Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Nov 5.
Published in final edited form as: Prev Med. 2017 Jun 7;101:149–155. doi: 10.1016/j.ypmed.2017.06.007

The role of neighborhood economic context on physical activity among children: Evidence from the Geographic Research onWellbeing (GROW) study

Yeonwoo Kim a,*, Catherine Cubbin a,b
PMCID: PMC10627422  NIHMSID: NIHMS1913109  PMID: 28601623

Abstract

Less than half of young children in the U.S. meet physical activity recommendations. While neighborhood economic context has been shown to be associated with physical activity, it is unknown whether this association varies according to family economic context. This study thus investigates whether neighborhood economic context, measured by poverty concentration and income inequality, are associated with physical activity among poor and non-poor children using data from the Geographic Research on Wellbeing study, California, 2012–2013 (N = 2670). Poor children who resided in (a) poor and equal neighborhoods or (b) non-poor and equal neighborhoods were more likely to engage in sufficient physical activity than were children residing in non-poor and unequal neighborhoods. Poor children in (a) non-poor and equal neighborhoods, (b) poor and equal neighborhoods, or (c) poor and unequal neighborhoods were less likely to report insufficient physical activity than those in non-poor and unequal neighborhoods. Neighborhood economic context was not associated with physical activity among non-poor children. Findings suggest that neighborhood economic context presents a social barrier to physical activity among poor children. Increasing physical activity among poor children in non-poor and unequal neighborhoods should be a high policy priority. Perceived social cohesion, perceived neighborhood safety, and park and walkability indicators did not mediate the associations between neighborhood economic context and physical activity. Further research needs to explore the mechanisms by which neighborhood economic context affects physical activity among children.

Keywords: Neighborhood economic context, Neighborhood poverty, Neighborhood income inequality, Physical activity

Physical inactivity is a factor for child health outcomes including obesity, hypercholesterolemia, and cardiovascular disease (Janssen and LeBlanc, 2010). Physical activity is also important from a psychosocial perspective: physical activity contributes to moral and social development, positive self-perceptions, and commitment to active living among children (Smith, 2003). However, when being objectively measured using accelerometers, only 42% of children aged 6–11 years in the United States meet physical activity recommendations (Troiano et al., 2008). The prevalence of physical activity has remained low during the last two decades (for review, Booth et al., 2015; Ekelund et al., 2011).

Neighborhood economic context is a plausible factor that may influence physical activity among children. According to the ecological framework, which considers individuals in the context of ecological systems where they reside, neighborhood characteristics influence children living within them (Leventhal and Brooks-Gunn, 2000). Neighborhood economic context can be investigated in two ways: absolute and relative economic context – which parallels the distinction drawn between absolute and relative poverty (Aber et al., 1997). Because neighborhood poverty does not assess relative economic context in a neighborhood – i.e., the distribution of economic resources or opportunities within a neighborhood – concurrent analysis of neighborhood poverty (indicating absolute economic context) and income inequality (indicating relative economic context) is worthy of exploration. The present study thus examines the association of both neighborhood poverty and income inequality with physical activity in a sample of children aged 4–10 years old across family poverty status.

1. Neighborhood poverty and child physical activity

There are different hypotheses to explain the effects of neighborhood poverty on physical activity among children. First, neighborhoods with a high concentration of poverty (hereafter, poor neighborhoods) may negatively influence children’s physical activity through lack of healthy norms that promote physical activity or through low quality of resources–such as playgrounds, police protection and schools–which can promote children’s physical activity (Jencks and Mayer, 1990). However, living in poor neighborhoods may not negatively influence non-poor children because non-poor children and their families are able to find high-quality services and better social support out of the poor neighborhood by using their financial resources. Second, the presence of economically advantaged neighbors may negatively affect poor children’s physical activity because poor children in neighborhoods with a low concentration of poverty (hereafter, non-poor neighborhoods) may feel relatively deprived while recognizing the gap between the economic status of the poor child’s family and the overall non-poor economic status of the neighborhood (Jencks and Mayer, 1990). One conceptual study suggests that the negative association between relative deprivation and individual health is mediated by psychological distress (Marmot and Wilkinson, 2001). Given empirical evidence suggesting a relationship between psychosocial variables, such as depression, and physical activity in children (Motl et al., 2004), it may be expected that poor children in non-poor neighborhoods feel relatively deprived, experience psychological distress, and are thus less likely to participate in physical activity. Furthermore, poor children in non-poor neighborhoods could have fewer peers of similar economic status compared to those in poor neighborhoods. Children are more likely to socialize with those from the same economic class (Weinger, 2000), which may lead to an exclusion from sports teams or other fee-based physical activities for poor children in non-poor neighborhoods. In summary, non-poor children may benefit from living in non-poor neighborhoods and avoid the harmful impact of living in poor neighborhoods by utilizing family-level financial resources. On the contrary, poor children may not be able to take advantage of better resources and healthy social norms in non-poor neighborhoods because of relative deprivation and may not be able to avoid the negative impact of living in poor neighborhoods due to limited financial resources.

Previous literature has shown mixed results for the relationship between neighborhood economic context and physical activity in children; for example, while economically disadvantaged neighborhoods were associated with less physical activity among U.S. children at fifth grade in one study (Pabayo et al., 2011), other studies reported no association with physical activity among young or poor children (Lovasi et al., 2011; Salomonsen-Sautel, 2011; Voorhees et al., 2009). Specifically, neighborhood poverty was not associated with physical activity during both warmer and cooler seasons among poor preschoolers in New York City (Lovasi et al., 2011) and with physical activity among children aged 5–14 years in Denver (Salomonsen-Sautel, 2011). Voorhees et al. (2009) measured neighborhood economic context by the Townsend index (Phillimore et al., 1994) and found no association with physical activity among sixth-grade girls in the U.S. These inconsistent results may reflect a lack of considering potential differential effects of neighborhood poverty by family poverty status, relative economic context in neighborhoods, or both. Analysis of both neighborhood poverty and income inequality separately by family poverty status may provide a clearer explanation of the association between neighborhood economic context and child physical activity.

2. Neighborhood income inequality and child physical activity

Income inequality is another potential indicator of neighborhood economic context influencing child physical activity. A high level of income inequality increases socioeconomic status-related anxiety, which may increase psychological distress and negatively affect health as a result (Wilkinson and Pickett, 2006). Additionally, a high level of income inequality is associated with lower levels of social trust (Blau and Blau, 1982), social support (Kawachi et al., 1999), social integration (Tumin, 1953), and investment in social infrastructure and health services (Ben-Shlomo et al., 1996; Smith, 1996). Because empirical evidence suggests that low levels of social trust, social support, and investment in social infrastructure are associated with a low level of children’s physical activity (Franzini et al., 2009; Gordon-Larsen et al., 2006), it is expected that neighborhood income inequality is related to psychological distress and low levels of social trust, social support, social integration, and investment in social infrastructure, which may discourage physical activity in young children.

The importance of neighborhood income inequality for health outcomes — for example, perceived health status, birth outcomes, and morbidity — has been examined in literature specific to adults (e.g., Massa et al., 2016; Nkansah-Amankra et al., 2010; Wen et al., 2003). Another study found that higher regional income inequality was associated with worse dental health among Brazilian children (Pattussi et al., 2001); however, it is unknown whether an inequality effect at the neighborhood level applies to children’s health behaviors, specifically physical activity.

In response to the compelling need to identify environmental obstacles that may be targeted in social policies to support physical development of young children, the present study examines the relationship between neighborhood economic context (i.e., neighborhood poverty and income inequality) and physical activity separately for poor children and non-poor children. Additionally, previous literature pointed out neighborhood social cohesion, neighborhood safety, and built environments as potential mediators in the association between neighborhood economic context and child physical activity (Babey et al., 2015; Carroll-Scott et al., 2013; Fueyo et al., 2016; Lovasi et al., 2011). The present study thus tests the mediating role of these factors in the relationship between neighborhood economic context and child physical activity.

3. Methods

3.1. Data

The Geographic Research on Wellbeing (GROW) study (2012−2013) is a population-based study of mothers in California designed to examine neighborhood effects on behavioral risk factors for cancer among women and their children. GROW is a follow-up survey of participants in the Maternal and Infant Health Assessment (MIHA), an annual, state-wide representative survey of mothers who gave birth to a live infant in California by mail or phone. GROW was collected from 3016 mothers. The GROW study was approved by the Institutional Review Boards at the University of Texas at Austin, the University of California, San Francisco, and the California Department of Public Health. Additional details about the MIHA and GROW surveys have been reported elsewhere (Cubbin, 2015; Cubbin et al., 2002). For information on neighborhood economic context, the final dataset was linked to the American Community Survey 2005–2009 via census geocodes based on the respondent’s address at the time of GROW.

The analytic sample for the current study included women living in California who responded for the index child based on reported birth date (excluding 147 mothers). We then included women who had non-missing data on child’s physical activity (excluding 161 mothers) and on marital status, maternal education level, and neighborhood economic context (excluding 38 mothers). The final analytic sample size thus is 2670 (718 poor and 1952 non-poor children).

3.2. Measures

The dependent variables were 1) sufficient physical activity and 2) insufficient physical activity among index children. Sufficient physical activity was defined as yes if a mother reported that her child engaged in physical activity on 5 days or more per week (otherwise no). Insufficient physical activity was defined as yes if a mother reported that her child engaged in physical activity on 2 days or less per week (otherwise no).

Multiple imputation was conducted for missing income values (about 9% of the overall sample) using hot-deck methodology and the following variables: age, race/ethnicity, education, employment status, marital status and neighborhood poverty. Poor women were defined as those with incomes at or below 100% of the poverty level and non-poor women were defined as those with incomes above 100% of the federal poverty level.

Neighborhood economic context was based on a combination of neighborhood poverty concentration (proportion of families with income that was below the federal poverty level for the census tract in which the respondent resided) and neighborhood income inequality (GINI coefficient at the census tract level). The GINI coefficient summarizes the overall level of inequality ranging between 0 (perfect equality) and 1 (perfect inequality; Sastry and Pebley, 2010). Neighborhoods were categorized into non-poor neighborhoods (census tracts below median poverty rate; <10.2%) and poor neighborhoods (census tracts at or above median poverty rate; ≥10.2%). Likewise, neighborhoods were categorized into equal neighborhoods (census tracts below the median GINI index; <0.35) and unequal neighborhoods (census tracts at or above the median GINI index; ≥0.35). The independent variable was neighborhood economic context (4 categories: non-poor and equal neighborhoods; non-poor and unequal neighborhoods; poor and equal neighborhoods; and poor and unequal neighborhoods). Considering poor children in non-poor and unequal neighborhoods may have multiple vulnerabilities, i.e., 1) feeling relatively deprived and with fewer peers of similar poverty status; and 2) low levels of social trust, support, integration, and resources in their neighborhoods, this study used neighborhoods with non-poor and unequal neighborhoods as the reference.

Mediators included five variables: 1) mothers’ perception of social cohesion; 2) mothers’ perception of neighborhood safety, 3) objectively measured straight line distance to the closest park (meters), 4) objectively measured park acreage within 0.5 mile of the child’s residence, and 5) objectively measured intersections per square mile within 0.5 mile of the child’s residence. For perceived social cohesion, we used a modified five-item scale (e.g., “my neighbors feel connected to each other.”) that Sampson et al. (1997) developed. The scale was based on four response options (strongly agree to strongly disagree; Cronbach’s alpha = 0.90) and the five items were then summated for a perceived social cohesion scale (range 5–20). Perceived neighborhood safety consisted of three questions about safety in daytime, in nighttime, and from crime (e.g., “I feel comfortable going to the park or playground closest to where I live during the day,”) based on four response options (strongly agree to strongly disagree). The three items were then summated to a perceived neighborhood safety scale (range 3–12). For built environments, we included three variables: straight line distance to the closest park (meters), park acreage within 0.5 mile from the child’s residence, and intersections per square mile within 0.5 mile from the child’s residence, as indicators of park availability and neighborhood walkability. Other variables included child’s age, gender, and race/ethnicity, and mother’s marital status, and educational attainment.

3.3. Analysis

We first examined the distribution of all variables overall, and by the prevalence of physical activity. Next, the proposed mediators were examined by neighborhood economic context. Third, we tested the effect of neighborhood economic context on each physical activity outcome (sufficient and insufficient physical activity). Because 83–92% of the participants were the only or one of only two participants in their census tract for each category of family poverty status, clustering within census tracts was not extensive; thus, we did not use multi-level modeling. Because the prevalence of sufficient or insufficient physical activity is >10%, the estimated odds ratio can overestimate risk ratio (Zhao, 2013). This study thus used a series of log-binomial regression models, estimated by the SAS GENMOD procedure with the WEIGHT statement and the REPEATED statement to specify the original weight, stratum, and cluster variables for accommodating correlated data (Hale et al., 2013), to estimate risk ratios (RR): (model 1) an unadjusted model; and (model 2) a ‘neighborhood’ model, which then adds the neighborhood economic context variable with control variables. Fourth, we tested proposed mediating effects in the relationship between neighborhood economic context and physical activity according to Baron and Kenny (1986). Potential mediators were entered into the multivariate models one at a time. As sensitivity analyses, we performed logistic regression models, estimated with the SAS SURVEYLOGISTIC procedure, to compare results obtained with SAS GENMOD. All analyses were performed separately for poor and non-poor children.

4. Results

Table 1 presents distributions of the variables and prevalence of physical activity by family poverty status. About a quarter were pre-schoolers, and three quarters of the sample were young children (ages 6–10). Over half of the children were boys, and a half of the children were Latino/a, followed by White and Black. Among poor children, one half of the children engaged in physical activity on 5 days or more per week (i.e., sufficient activity), and 25% of children participated in physical activity only on 2 days or less per week (i.e., insufficient activity). In general, poor children had higher levels of physical activity than non-poor children. However, poor children showed higher variability in prevalence of physical activity by type of neighborhood economic context (sufficient activity: 35–55%; insufficient activity; 20–39%) compared with non-poor children (sufficient activity: 33–46%; insufficient activity: 25–34%).

Table 1.

Descriptive statistics and prevalence of physical activity by individual and neighborhood characteristics (weighted), Geographic Research on Wellbeing (GROW) Study, California, US, 2012–2013, N = 2670.

Characteristic Total (%) Prevalence of physical activity per week (%)
Poor children Non-poor children
≥5 days ≤2 days ≥5 days ≤2 days
Total 100.0 49.5 24.6 40.9 28.4
Child’s age
 4–5 years 22.0 57.9 18.2 42.3 32.8
 6–7 years 37.9 46.3 25.7 42.3 28.5
 8–10 years 40.2 46.6 27.9 38.8 26.3
Child’s gender
 Boy 51.4 52.3 20.5 42.1 26.7
 Girl 48.6 46.4 29.2 39.6 30.3
Child’s race/ethnicity
 Black 4.9 47.5 24.1 38.8 30.8
 Latino/a 52.4 49.1 25.0 40.2 26.1
 White 30.1 50.7 23.0 41.5 31.1
 Other/missing 7.6 60.2 19.1 42.1 26.9
Mother’s marital status
 Married or living together 83.9 48.1 26.9 40.7 28.8
 Unmarried 16.1 53.5 18.1 42.3 26.1
Mother’s education
 Less than high school 11.1 47.0 34.2 45.7 32.9
 High-school graduate/GED 32.4 49.6 21.3 40.2 26.0
 Some college or above 56.5 52.1 21.8 40.8 28.9
Family income (% of federal poverty level)
 ≤100% 32.2 49.5 24.6
 101–200% 20.4 41.3 27.2
 201–300% 10.7 41.0 31.2
 301–400% 7.9 36.5 28.2
 >400% 28.9 41.7 28.4
Neighborhood economic context
 Non-poor & equal 40.2 53.9 21.5 39.6 29.9
 Non-poor & unequal 7.4 35.0 38.8 42.2 25.9
 Poor & equal 11.9 54.5 20.4 33.1 33.9
 Poor & unequal 40.4 48.6 25.0 45.6 24.7
Open in a new tab

Neighborhood economic context varied both by family poverty status and neighborhood characteristics conceptualized as potential mediators (Fig. 1). While thirds of poor children lived in poor and unequal neighborhoods, most non-poor children lived in non-poor and equal neighborhoods. Non-poor and equal neighborhoods tended to have the most health-promoting social and built environments compared to other neighborhood economic context types, although the differences were not great. On the other hand, within each neighborhood economic context type, non-poor residents perceived neighborhoods as more cohesive and safer and had “healthier” built environments than poor residents. This trend was especially notable in non-poor neighborhoods.

Fig. 1.

Fig. 1.

Open in a new tab

Children’s neighborhood characteristics by neighborhood economic context and family economic status, Geographic Research on Wellbeing (GROW) Study, California, US, 2012–2013, N = 2670.

Table 2 summarizes the results of the regression models of physical activity among children separately by family poverty status. As shown in Panel A, Model 2, poor children who resided in non-poor and equal or poor and equal neighborhoods were more likely to engage in sufficient activity than were children residing in non-poor and unequal neighborhoods. In Panel B, Model 2, poor children who resided in non-poor and unequal neighborhoods had higher risk of insufficient activity than those in other neighborhood types. Among non-poor children, neighborhood economic context was not associated with physical activity. We reanalyzed the data using the SAS SURVEYLOGISTIC procedure to account for the complex survey design and found similar results.

Table 2.

Regression analysis assessing associations between neighborhood economic context and physical activity by individual poverty status, Geographic Research on Wellbeing (GROW) Study, California, US, 2012–2013, N = 2670.

Characteristic Poor children Non-poor children
Model 1 Model 2 Model 1 Model 2
RR 95% CI RR 95% CI RR 95% CI RR 95% CI
A. Sufficient physical activity: ≥5 days per week (%)
Child’s age 0.94 0.90–0.99 0.95 0.90–0.99 0.97 0.93–1.01 0.97 0.93–1.01
Child’s gender
 Boy 1.13 0.96–1.32 1.06 0.91–1.23 1.06 0.94–1.20 1.06 0.94–1.19
 Girl 1.00 1.00 1.00 1.00
Child’s race/ethnicity
 Black 0.94 0.60–1.47 0.79 0.51–1.20 0.97 0.85–1.10 0.88 0.69–1.14
 Latino/a 0.97 0.67–1.41 0.88 0.64–1.21 0.94 0.73–1.19 0.94 0.81–1.09
 White 1.00 1.00 1.00 1.00
 Other and missing 1.19 0.72–1.95 1.07 0.69–1.65 1.02 0.82–1.25 1.00 0.81–1.24
Mother’s marital status
 Married or living together 0.90 0.75–1.07 0.88 0.74–1.04 0.96 0.80–1.15 0.95 0.79–1.14
 Unmarried 1.00 1.00 1.00 1.00
Mother’s education
 Less than high school 1.00 1.00 1.00 1.00
 High-school graduate/GED 1.06 0.86–1.29 1.02 0.84–1.23 0.88 0.65–1.20 0.83 0.62–1.11
 Some college or above 1.11 0.88–1.41 1.01 0.80–1.28 0.89 0.67–1.18 0.83 0.63–1.10
Neighborhood economic context
 Non-poor & equal 1.54 0.96–2.47 1.62 1.00–2.62 0.94 0.75–1.16 0.92 0.74–1.14
 Non-poor & unequal 1.00 1.00 1.00 1.00
 Poor & equal 1.56 0.98–2.49 1.62 1.00–2.62 0.78 0.58–1.05 0.77 0.57–1.04
 Poor & unequal 1.39 0.89–2.17 1.44 0.91–2.28 1.08 0.86–1.35 1.08 0.86–1.36
B. Insufficient physical activity: ≤2 days per week (%)
Child’s age 1.11 1.02–1.21 1.09 1.01–1.19 0.94 0.89–1.00 0.94 0.89–0.99
Child’s gender
 Boy 0.70 0.53–0.92 0.73 0.56–0.96 0.88 0.75–1.04 0.88 0.75–1.04
 Girl 1.00 1.00 1.00 1.00
Child’s race/ethnicity
 Black 1.05 0.46–2.38 1.35 0.63–2.90 0.99 0.74–1.33 1.07 0.79–1.43
 Latino/a 1.09 0.54–2.17 1.06 0.59–1.91 0.84 0.70–1.00 0.82 0.67–1.01
 White 1.00 1.00 1.00 1.00
 Other and missing 0.83 0.31–2.22 0.76 0.28–2.07 0.87 0.65–1.15 0.86 0.65–1.14
Mother’s marital status
 Married or living together 1.49 1.05–2.11 1.45 1.03–2.04 1.10 0.85–1.44 1.05 0.81–1.37
 Unmarried 1.00 1.00 1.00 1.00
Mother’s education
 Less than high school 1.00 1.00 1.00 1.00
 High-school graduate/GED 0.62 0.46–0.84 0.66 0.49–0.88 0.79 0.51–1.22 0.76 0.50–1.16
 Some college or above 0.64 0.43–0.94 0.71 0.47–1.08 0.88 0.59–1.30 0.78 0.52–1.17
Neighborhood economic context
 Non-poor & equal 0.55 0.32–0.96 0.59 0.36–0.98 1.15 0.84–1.58 1.17 0.85–1.61
 Non-poor & unequal 1.00 1.00 1.00 1.00
 Poor & equal 0.52 0.30–0.91 0.54 0.32–0.90 1.31 0.90–1.90 1.39 0.96–2.01
 Poor & unequal 0.64 0.42–1.00 0.66 0.45–0.95 0.95 0.68–1.34 1.00 0.71–1.41
Open in a new tab

RR = risk ratio. CI = confidence interval.

Finally, we tested five potential mediators among poor children. We first tested the association between neighborhood economic context and each potential mediator and found significant associations with all five mediators (results not shown). Subsequent separate models examined the effect of each potential mediator on the association between neighborhood economic context and physical activity among poor children (Table 3). Each potential mediator was added to the multivariate model one at a time. For example, multivariate model 2 is simply the risk ratios (RR = 1.44–1.62) of neighborhood economic context that was presented in Table 2, Panel A, Model 2 for poor children. Adding perceived social cohesion to that multivariate model, we observed that social cohesion was not associated with sufficient activity (RR = 1.01), and the risk ratios of physical activity associated with neighborhood economic context barely changed. The risk ratio for each potential mediator in an unadjusted model is also shown, to compare with the adjusted risk ratio. In all cases, there is no substantial evidence of mediation; that is, mediators were not associated with physical activity in unadjusted or adjusted models and, in each subsequent model, the risk ratios of physical activity did not significantly change when the mediator was included in the model.

Table 3.

Examining mediating effects onthe pathway between neighborhood economic context and physicalactivity, Geographic Research on Wellbeing (GROW) Study, California, US, 2012–2013.

Characteristic Mediator, unadjusted model RR (95% CI) Mediator, adjusted model RR (95% CI) Neighborhood economic context
Non-poor & equal RR (95% CI) Poor & equal RR (95% CI) Poor & unequal RR (95% CI)
A. Sufficient physical activity: ≥5 days per week (%)
Poor children, n = 718
Multivariate model 2 (from Table 3) 1.62 (1.00–2.62) 1.62 (1.00–2.62) 1.44 (0.91–2.28)
Perceived social cohesion 1.01 (0.98–1.04) 1.01 (0.99–1.04) 1.63 (0.93–2.84) 1.66 (0.95–2.90) 1.49 (0.87–2.55)
Perceived neighborhood safety 1.04 (1.00–1.09) 1.05 (1.00–1.09) 1.57 (0.96–2.56) 1.69 (1.05–2.73) 1.49 (0.93–2.38)
Distance to nearest park (per 250 m) 1.00 (0.96–1.07) 1.02 (0.97–1.08) 1.72 (1.01–2.91) 1.69 (1.00–2.86) 1.52 (0.91–2.51)
Park acreage within 0.5 mile (per 30 acres) 1.03 (0.94–1.12) 1.00 (0.91–1.09) 1.71 (1.01–2.91) 1.68 (0.99–2.84) 1.51 (0.91–2.50)
Intersections per square mile within 0.5 mile (per 20 intersections) 0.99 (0.96–1.02) 1.00 (0.97–1.02) 1.66 (0.99–2.77) 1.63 (0.98–2.72) 1.46 (0.90–2.38)
B. Insufficient physical activity: ≤2 days per week (%)
Poor children, n = 718
Multivariate model 2 (from Table 3) 0.59 (0.36–0.98) 0.54 (0.32–0.90) 0.66 (0.45–0.95)
Perceived social cohesion 0.97 (0.93–1.02) 0.97 (0.93–1.01) 0.69 (0.41–1.16) 0.49 (0.27–0.87) 0.71 (0.48–1.05)
Perceived neighborhood safety 0.94 (0.88–1.01) 0.95 (0.88–1.03) 0.68 (0.39–1.18) 0.54 (0.31–0.95) 0.68 (0.44–1.05)
Distance to nearest park (per 250 m) 1.04 (0.96–1.14) 1.02 (0.93–1.12) 0.61 (0.36–1.03) 0.52 (0.30–0.90) 0.66 (0.45–0.97)
Park acreage within 0.5 mile (per 30 acres) 0.84 (0.64–1.10) 0.89 (0.69–1.16) 0.61 (0.37–1.03) 0.52 (0.30–0.90) 0.65 (0.44–0.95)
Intersections per square mile within 0.5 mile (per 20 intersections) 1.01 (0.95–1.06) 1.00 (0.95–1.06) 0.60 (0.36–1.01) 0.51 (0.30–0.89) 0.64 (0.44–0.94)
Open in a new tab

5. Discussion

In this study, our results suggest that living in a non-poor and unequal neighborhood is more harmful for physical activity among poor young children compared with other neighborhood economic context types. On the other hand, there was no association between neighborhood economic context and physical activity among non-poor young children. To our knowledge, this is the first study to concurrently link neighborhood poverty and income inequality to child physical activity and examine associations separately by family economic status.

A few studies have examined childhood physical activity and neighborhood economic contexts, including neighborhood poverty rate (Lovasi et al., 2011; Salomonsen-Sautel, 2011), median family income (Salomonsen-Sautel, 2011), and a composite index combining neighborhood poverty rate and other economic indicators (Leung et al., 2010; Pabayo et al., 2011). However, only one study found a significant association between neighborhood economic context and child physical activity (Pabayo et al., 2011). The current study suggests that the mostly insignificant results in previous literature may be because of not considering the differential effects of neighborhood poverty by family-level economic status, the relative economic context in neighborhoods, or both. Neighborhood poverty could be differentially associated with child physical activity, depending on neighborhood income inequality. Residents in unequal neighborhoods may experience socioeconomic status-related anxiety and low levels of social trust, support, integration, and resources in neighborhoods compared to neighborhoods with more income equality (Ben-Shlomo et al., 1996; Blau and Blau, 1982; Kawachi et al., 1999; Tumin, 1953; Smith, 1996; Wilkinson and Pickett, 2006). The impact of high inequality may be more harmful for poor children in non-poor neighborhoods than those in poor neighborhoods due to feeling relatively deprived and having fewer peers of similar poverty status. This is suggested by the present study: non-poor and unequal neighborhoods were associated with about 40% increase in risk of insufficient activity among poor children (compared with poor children in other types of neighborhoods). While non-poor neighborhoods are considered to have more advantaged “healthy” neighbors and better resources, such as playgrounds, better police protection, and quality of teachers (Jencks and Mayer, 1990), the current study found that such presumably healthy environments (socially cohesive and safe neighborhoods and good built environments) may not benefit poor children living in them (vs. non-poor children in the same type of neighborhoods) (see Fig. 1). These results suggest that poor children in non-poor and unequal neighborhoods could suffer from a double impact: 1) feeling relatively deprived and having fewer peers of similar poverty status; and 2) low levels of social trust, support, integration, and resources in neighborhoods. On the other hand, neighborhood poverty and income inequality had no influence on physical activity of non-poor children perhaps because they may have access to family financial resources to avoid the harmful impact of neighborhood economic disadvantage – such as by utilizing better physical activity facilities and services outside of the neighborhoods.

In contrast to theoretical perspectives, perceived social cohesion, perceived neighborhood safety, and built environment characteristics did not significantly mediate the associations between neighborhood economic context and physical activity in the present study, and bivariate correlations were low between neighborhood economic context and the proposed mediators (although significant, r = 0.06 to 0.25). Cross-sectional analyses are not ideal to examine mediating effects because of uncertain temporal sequencing and lagged effects of neighborhood characteristics on health (Alvarado, 2011). Future research thus needs to use longitudinal data and to investigate long-term effects of economic, physical, and social neighborhoods and their direct and indirect impacts on physical activity.

Several limitations deserve mention. Neighborhoods are based on census tracts, comprising about 4000 individuals. Census tract boundaries, although designed to capture a homogenous geographic area with visible boundaries and residents of similar socio-demographic characteristics (Eschbach et al., 2004; Ross and Mirowsky, 2008), are still administrative and artificial boundaries (Voorhees et al., 2009). As well, because mobility in the United States is high and neighborhood economic context changes over time (Margerison-Zilko et al., 2015), the census tract at the time of survey may not represent the same type of neighborhood throughout childhood. In terms of measurement, we measured a single, mother-reported item on her child’s physical activity. The mother-reported measure of physical activity is prone to recall bias and social desirability (Klesges et al., 2004). This measure does not capture specific contexts of physical activity – such as active transport, organized sports, and physical education at school. Future research should examine other physical activity domains and measure physical activity objectively (such as via accelerometers). We chose to focus on absolute and relative economic indicators of neighborhood economic context, but other important neighborhood demographic and economic measures – such as racial/ethnic composition, employment patterns, housing status, and assets – should be considered in future research. Furthermore, longitudinal data would strengthen causal inferences by which neighborhood economic context – or its trajectories- affect child physical activity. To increase our understanding of neighborhood economic context effects, further studies are needed to elucidate the mechanisms by which neighborhood economic context influences child health behaviors.

Despite these limitations, this study has several strengths. First, the data represents a large, ethnically diverse sample. In addition, rather than relying only upon an absolute economic indicator of neighborhood economic context, we calculated neighborhood income inequality as well, based on knowledge suggesting that absolute and relative economic measures of neighborhoods should be concurrently taken into account. The main strength of our study is the examination of vulnerability to health behavior among poor children residing in non-poor and unequal neighborhoods. Although the percent of poor children living in non-poor and unequal neighborhoods was small, the public health significance is large because of the serious potential consequences of insufficient physical activity on child health. In order to encourage physical activity among poor children in non-poor and unequal neighborhoods, investment in healthy built and social environments are needed. These results also suggest that mobility programs to move low-income families from poor neighborhoods to non-poor neighborhoods, such as Moving to Opportunity (Sanbonmatsu et al., 2011), needs to offer mobility choices that consider both neighborhood poverty and income inequality.

Acknowledgement

This work was supported by a Research Scholar Grant from the American Cancer Society (C.C., grant number RSGT-11-010-01-CPPB). The American Cancer Society had no role in the design, analysis or writing of this article.

References

  1. Aber JL, Bennett NG, Conley DC, Li J, 1997. The effects of poverty on child health and development. Annu. Rev. Public Health 18 (1), 463–483. [DOI] [PubMed] [Google Scholar]
  2. Alvarado SE, 2011. The Effect of Neighborhood Context on Obesity and Educational Achievement Among Youth. University of Wisconsin – Madison. [Google Scholar]
  3. Babey SH, Tan D, Wolstein J, Diamant AL, 2015. Neighborhood, family and individual characteristics related to adolescent park-based physical activity. Prev. Med. 76, 31–36. [DOI] [PubMed] [Google Scholar]
  4. Baron RM, Kenny DA, 1986. The moderator–mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J. Pers. Soc. Psychol. 51 (6), 1173–1182. [DOI] [PubMed] [Google Scholar]
  5. Ben-Shlomo Y, White IR, Marmot M, 1996. Does the variation in the socioeconomic characteristics of an area affect mortality? BMJ 312 (7037), 1013–1014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Blau JR, Blau PM, 1982. The cost of inequality: metropolitan structure and violent crime. Am. Sociol. Rev. 47 (1), 114–129. [Google Scholar]
  7. Booth VM, Rowlands AV, Dollman J, 2015. Physical activity temporal trends among children and adolescents. J. Sci. Med. Sport 18 (4), 418–425. [DOI] [PubMed] [Google Scholar]
  8. Carroll-Scott A, Gilstad-Hayden K, Rosenthal L, Peters SM, McCaslin C, Joyce R, Ickovics JR, 2013. Disentangling neighborhood contextual associations with child body mass index, diet, and physical activity: the role of built, socioeconomic, and social environments. Soc. Sci. Med. 95, 106–114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Cubbin C, 2015. Survey methodology of the geographic research on wellbeing (GROW) study. BMC 8 (1), 1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cubbin C, Braveman PA, Marchi KS, Chavez GF, Santelli JS, Gilbert BJC, 2002. Socioeconomic and racial/ethnic disparities in unintended pregnancy among postpartum women in California. Matern. Child Hlth. J. 6 (4), 237–246. [DOI] [PubMed] [Google Scholar]
  11. Ekelund U, Tomkinson G, Armstrong N, 2011. What proportion of youth are physically active? Measurement issues, levels and recent time trends. Br. J. Sports Med. 45 (11), 859–865. [DOI] [PubMed] [Google Scholar]
  12. Eschbach K, Ostir GV, Patel KV, Markides KS, Goodwin JS, 2004. Neighborhood context and mortality among older Mexican Americans: is there a barrio advantage? Am. J. Public Health 94 (10), 1807–1812. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Franzini L, Elliott MN, Cuccaro P, Schuster M, Gilliland MJ, Grunbaum JA, … Tortolero SR, 2009. Influences of physical and social neighborhood environments on children’s physical activity and obesity. Am. J. Public Health 99 (2), 271–278. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Fueyo JL, Garcia LMT, Mamondi V, Alencar GP, Florindo AA, Berra S, 2016. Neighborhood and family perceived environments associated with children’s physical activity and body mass index. Prev. Med. 82, 35–41. [DOI] [PubMed] [Google Scholar]
  15. Gordon-Larsen P, Nelson MC, Page P, Popkin BM, 2006. Inequality in the built environment underlies key health disparities in physical activity and obesity. Pediatrics 117 (2), 417–424. [DOI] [PubMed] [Google Scholar]
  16. Hale JJ, Thompson MD, Darden MP, 2013. Calculating subset weighted analysis using PROC SURVEYFREQ and GENMOD. SAS Global Forum 2013. [Google Scholar]
  17. Janssen I, LeBlanc AG, 2010. Systematic review of the health benefits of physical activity and fitness in school-aged children and youth. Int. J. Behav. Nutr. Phy. 7 (40), 1–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Jencks C, Mayer SE, 1990. The social consequences of growing up in a poor neighborhood. In: Lynn L Jr, McGeary MGH (Eds.), Inner-city Poverty in the United States. National Academy Press, Washington, DC, pp. 111–186. [Google Scholar]
  19. Kawachi I, Kennedy BP, Wilkinson RG, 1999. Crime: social disorganization and relative deprivation. Soc. Sci. Med. 48 (6), 719–731. [DOI] [PubMed] [Google Scholar]
  20. Klesges LM, Baranowski T, Beech B, Cullen K, Murray DM, Rochon J, Pratt C, 2004. Social desirability bias in self-reported dietary, physical activity and weight concerns measures in 8-to 10-year-old African-American girls: results from the Girls Health Enrichment Multisite Studies (GEMS). Prev. Med. 38, S78–S87. [DOI] [PubMed] [Google Scholar]
  21. Leung CW, Gregorich SE, Laraia BA, Kushi LH, Yen IH, 2010. Measuring the neighborhood environment: associations with young girls’ energy intake and expenditure in a cross-sectional study. Int. J. Behav. Nutr. Phy. 7 (1), 1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Leventhal T, Brooks-Gunn J, 2000. The neighborhoods they live in: the effects of neighborhood residence on child and adolescent outcomes. Psychol. Bull. 126 (2), 309–337. [DOI] [PubMed] [Google Scholar]
  23. Lovasi GS, Jacobson JS, Quinn JW, Neckerman KM, Ashby-Thompson MN, Rundle A, 2011. Is the environment near home and school associated with physical activity and adiposity of urban preschool children? J. Urban Health 88 (6), 1143–1157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Margerison-Zilko C, Cubbin C, Jun J, Marchi K, Fingar K, Braveman P, 2015. Beyond the cross-sectional: neighborhood poverty histories and preterm birth. Am. J. Public Health 105 (6), 1174–1180. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Marmot M, Wilkinson RG, 2001. Psychosocial and material pathways in the relation between income and health: a response to Lynch et al et al. Br. Med. J. 322 (7296), 1233–1236. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Massa KHC, Pabayo R, Lebrão ML, Chiavegatto Filho ADP, 2016. Environmental factors and cardiovascular diseases: the association of income inequality and green spaces in elderly residents of São Paulo, Brazil. BMJ Open 6 (9), e011850. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Motl RW, Birnbaum AS, Kubik MY, Dishman RK, 2004. Naturally occurring changes in physical activity are inversely related to depressive symptoms during early adolescence. Psychosom. Med. 66 (3), 336–342. [DOI] [PubMed] [Google Scholar]
  28. Nkansah-Amankra S, Dhawain A, Hussey JR, Luchok KJ, 2010. Maternal social support and neighborhood income inequality as predictors of low birth weight and preterm birth outcome disparities: analysis of South Carolina Pregnancy Risk Assessment and Monitoring System survey, 2000–2003. Matern. Child Hlth. J. 14 (5), 774–785. [DOI] [PubMed] [Google Scholar]
  29. Pabayo R, Belsky J, Gauvin L, Curtis S, 2011. Do area characteristics predict change in moderate-to-vigorous physical activity from ages 11 to 15 years? Soc. Sci. Med. 72 (3), 430–438. [DOI] [PubMed] [Google Scholar]
  30. Pattussi MP, Marcenes W, Croucher R, Sheiham A, 2001. Social deprivation, income inequality, social cohesion and dental caries in Brazilian school children. Soc. Sci. Med. 53 (7), 915–925. [DOI] [PubMed] [Google Scholar]
  31. Phillimore P, Beattie A, Townsend P, 1994. Widening inequality of health in northern England, 1981–91. BMJ 308 (6937), 1125–1128. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Ross CE, Mirowsky J, 2008. Neighborhood socioeconomic status and health: context or composition? City Community 7 (2), 163–179. [Google Scholar]
  33. Salomonsen-Sautel S, 2011. A Social Ecological Study Examining Individual, Household, and Neighborhood Influences on Children’s Physical Activity and Active Commuting to/from School. University of Colorado at Denver. [Google Scholar]
  34. Sampson RJ, Raudenbush SW, Earls F, 1997. Neighborhoods and violent crime: a multilevel study of collective efficacy. Science 277 (5328), 918–924. [DOI] [PubMed] [Google Scholar]
  35. Sanbonmatsu L, Ludwig J, Katz LF, Gennetian LA, Duncan GJ, Kessler RC, … Lindau ST, 2011. Moving to Opportunity for Fair Housing Demonstration Program: Final Impacts Evaluation. U.S. Department of Housing and Urban development, Office of Policy Development and Research. [Google Scholar]
  36. Sastry N, Pebley AR, 2010. Family and neighborhood sources of socioeconomic inequality in children’s achievement. Demography 47 (3), 777–800. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Smith GD, 1996. Income inequality and mortality: why are they related? BMJ 312 (7037), 987–988. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Smith AL, 2003. Peer relationships in physical activity contexts: a road less traveled in youth sport and exercise psychology research. Psychol. Sport Exerc. 4 (1), 25–39. [Google Scholar]
  39. Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, McDowell M, 2008. Physical activity in the United States measured by accelerometer. Med. Sci. Sports Exerc. 40 (1), 181–188. [DOI] [PubMed] [Google Scholar]
  40. Tumin MM, 1953. Some principles of stratification: a critical analysis. Am. Sociol. Rev. 18 (4), 387–394. [Google Scholar]
  41. Voorhees CC, Catellier DJ, Ashwood JS, Cohen DA, Rung A, Lytle L, … Dowda M, 2009. Neighborhood socioeconomic status and non school physical activity and body mass index in adolescent girls. J. Phys. Act. Health 6 (6), 731–740. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Weinger S, 2000. Economic status: middle class and poor children’s views. Child. Soc. 14 (2), 135–146. [Google Scholar]
  43. Wen M, Browning CR, Cagney KA, 2003. Poverty, affluence, and income inequality: neighborhood economic structure and its implications for health. Soc. Sci. Med. 57 (5), 843–860. [DOI] [PubMed] [Google Scholar]
  44. Wilkinson RG, Pickett KE, 2006. Income inequality and population health: a review and explanation of the evidence. Soc. Sci. Med. 62 (7), 1768–1784. [DOI] [PubMed] [Google Scholar]
  45. Zhao K, 2013. Proper estimation of relative risk using PROC GENMOD in population studies. Western Users of SAS Software Conference. [Google Scholar]

RESOURCES