Socioeconomic Differences in Access to Neighborhood and Network Social Capital and Associations With Body Mass Index Among Black Americans

Abstract

Purpose:

To examine associations between socioeconomic status and two forms of social capital, namely, neighborhood and network measures, and how these distinct forms of capital are associated with body mass index (BMI) among Black residents of low-income communities.

Design:

Respondent-driven sampling was used to engage residents in a household survey to collect data on the respondents’ personal network, perceptions about their neighborhood environment, and health.

Setting:

Eight special emphasis neighborhoods in Greenville, South Carolina.

Participants:

N = 337 black/African American older adults, nearly half of whom have a household income of less than $15 000 and a high school education, were included.

Measures:

Neighborhood capital was assessed via three scales on social cohesion, collective efficacy, and social support from neighbors. Network capital was calculated via a position generator, common in egocentric network surveys. Body mass index was calculated with self-reported height and weight.

Analysis:

Multilevel linear regression models were used to examine the association between neighborhood and network capital and obesity among respondents within sampling chains.

Results:

Higher household income was associated with greater neighborhood capital, whereas higher educational attainment was associated with greater network capital. Social cohesion was negatively associated with BMI (b = −1.25, 95% confidence interval [CI]: −2.39 to −0.11); network diversity was positively associated with BMI (b = 0.31, 95% CI: 0.08 to 0.55).

Conclusion:

The findings shed light on how social capital may be patterned by socioeconomic status and, further, how distinct forms of capital may be differentially associated with health among black Americans.

Keywords

neighborhood social capital network diversity body mass index black Americans

Purpose

For over two decades, public health scientists have been increasingly interested in the role of social capital on health.¹ Broadly defined, social capital refers to resources that are accessible through social relationships,² including social support, material support, and information. Social capital is influential for multiple health outcomes, including mortality,³ mental health,⁴ and overall well-being.⁵ There is also a growing interest in the relationship between social capital and chronic conditions, including overweight and obesity.^6
-8 Suglia and colleagues suggest social capital may be linked to obesity through a broad range of mechanisms, from neighborhood characteristics, including racial diversity, to social norms for health behaviors, such as diet and exercise. Previous studies examining social capital and obesity have found that higher levels of social capital are associated with lower body mass index (BMI) and smaller waist circumference,^7,9 indicating a protective effect of social capital on weight status.

Empirical studies often measure single forms of social capital, despite its multidimensionality. Indeed, most studies that examine social capital and health have investigated community, or neighborhood-based aspects of social capital, which focus on group dynamics and engagement, including social cohesion (ie, social bonds), reciprocity (ie, mutual exchange), and collective efficacy (ie, ability of groups to enact change).^1,9,10 These constructs can be traced to Putnam et al’s work on civic communities,¹¹ Sampson and colleagues work on collective efficacy,^12,13 and a widely cited public health study that further delineated social capital as a community-level variable,¹⁴ with measures such as civic participation and levels of mutual trust among community members. However, the widespread adoption of this relatively narrow definition and subsequently singular approach to social capital measurement may limit the ability of public health scientists to more broadly conceptualize how social capital influences health.¹⁵ Another body of research has utilized network measures of social capital, which explore individual-level mechanisms linking social capital to health. While neighborhood approaches tend to focus on group processes (i.e., social cohesion of residents within a neighborhood), network capital focuses on the amount and quality of resources that an individual can access through specific relationships.¹⁶ Although researchers have articulated the importance of comparing as well as accounting for these two levels when studying the role of social capital on health,^17,18 few have done so. The current study will address this gap by examining the utility of measuring both neighborhood social capital and network diversity within a single sample and their relationship with BMI.

In a theoretical model put forth by Berkman and Glass,¹⁹ social relationships are thought to affect overall health through several pathways, which we draw upon in our study of social capital and BMI. Here, we postulate that neighborhood and network capital are similar but distinct constructs influencing broader health outcomes, including obesity. Specifically, social networks may be linked with obesity outcomes through the provision of financial and social support (buffering the effects of adversity and stress), information and decision-making regarding health care, and direct social influence and support for healthy behaviors (eg, diet and physical activity). Some scholars theorize that larger, more diverse networks not only increase the availability of support and variety of resources within the network²⁰ but may likely break up potentially harmful social and behavioral norms that occur among small and denser networks.¹⁶ Indeed, greater network diversity has been associated with greater self-rated health, lower physical inactivity, and lower risk of overweight and obesity.^7,16,21

On the other hand, neighborhood social capital is hypothesized to be associated with health and obesity via collective actions, including the enforcement of social, cultural, and moral norms.¹ Higher collective efficacy, for example, is hypothesized to reflect greater conformity and control of deviant behavior²² and may also promote healthy behaviors and foster a sense of belongingness. Neighborhood social capital is also indicative of greater political power for improving neighborhood physical and social conditions, including safety, green space, and litter, all which have been associated with physical activity^23
-25 as well as mental health and overall well-being.^26,27 Altogether, a growing body of literature underscores several indirect pathways by which neighborhood social capital may affect obesity,²⁸ including via stress and allostatic load, behavioral regulation, and walkable neighborhoods. Yet social capital has also been associated with poor health behaviors in some instances. Individuals with higher levels of both neighborhood and network social capital have also been shown to engage more often in problematic drinking^29,30 and smoking,^31,32 suggesting mechanisms of behavioral regulation and social control can also exert negative effects. Thus, while both forms of social capital are linked with health, the evidence remains mixed about whether these relationships are protective for overweight and obesity.

Despite the relevance of social capital for population health, limited research has explored the role of social capital on health outcomes, including obesity, within black communities.^33
-35 To that extent, it is not well understood whether or how social capital is associated with socioeconomic indicators or health outcomes among broader demographic groups. Compared to whites, black Americans have higher rates of chronic disease and mortality^36,37 and shorter life expectancy.³⁸ Given that black Americans are more likely to experience higher rates of obesity than whites or other racial and ethnic groups,^39,40 more studies are warranted that examine associations between social capital and BMI within this specific population.

Finally, while neighborhood capital is a collective asset that proposes to benefit members equally, network capital may not be evenly distributed across socioeconomic status.^19,41 Lin argues that inequality in access to network capital occurs through the socioeconomic positioning of specific social groups. That is, individuals with high levels of social capital have a greater ability to access and leverage network resources for health than do individuals with low levels of capital.^41
-43 Because social capital is based on the ability of individuals to connect with others through common links, such as mutual friendships, working at the same company, or living in close proximity to one another,^44,45 factors that structure these opportunities, such as educational attainment or income, may have important health ramifications.

In summary, the purpose of the current study is to explore associations between two key indicators of socioeconomic status and two forms of social capital, namely, neighborhood and network measures and to determine whether both forms of social capital are associated with BMI among black Americans.

Methods

Design

Respondent-driven sampling (RDS) was used to engage residents from each of the eight neighborhoods to participate in a household survey. The RDS includes a double incentive system that not only provides compensation to participants for completing the survey but also for successful recruitment of other participants. As such, new people are invited to participate via community members, rather than study personnel. These features allow a community to take ownership of the referral process and may make participation more inviting to those who are less likely to engage otherwise.^46,47

In the current study, neighborhood association presidents served as the “seed” in each neighborhood. Neighborhood presidents were asked to select ten residents of varying demographic characteristics (ie, gender, age, and occupation) who would serve as the initial (first) wave of the sampling chain. These ten people were given a coupon from the president which served as their invitation to the study and would also track how participants entered the study. After participants of the first wave completed the survey, they were asked to recruit three more individuals (a second wave) who lived in their neighborhood by giving them similar invitation/tracking coupons. All participants were given a $10 grocery store gift card for completing the survey and were incentivized to recruit other residents with the use of a raffle. Participants of the second wave were also asked to recruit three others, and so forth, for a total of four waves of participants. Specific to RDS methodology, the coupons contained identification numbers that linked participants with their recruiters and which were used to create sampling chains that were used as the cluster variable for multilevel analysis. A total of 180 sampling chains were created, ranging in size from 1 to 14 people across four waves, with an average of about two persons per chain.

Participants completed the survey at a community center or church located within their neighborhood. Eligibility for the survey included the ability to speak and comprehend English, being at least 18 years of age or older, noninstitutionalized, and residing in one of the eight study neighborhoods. This study obtained Institutional Review Board approval from the University of South Carolina and complied with APA ethical principles in the treatment of research participants. All participants provided written informed consent prior to beginning the survey.

Sample

The Greenville Healthy Neighborhoods Project occurred in 2014 in eight Special Emphasis neighborhoods located within the City of Greenville, South Carolina. The Special Emphasis designation represented a heightened effort on behalf of the City of Greenville to partner with disadvantaged communities in order to leverage existing resources and promote well-being among residents of those neighborhoods. Study neighborhoods represented a diverse mix of socioeconomic and demographic resident characteristics (Table 1). The majority of the neighborhoods were historically and predominantly black communities, ranging from 34% to 82% black residents. In all of the neighborhoods, more than 30% of the residents lived at or below the Federal Poverty Line, and annual household incomes averaged less than $18 000.⁴⁸

Table 1.

Sample Characteristics of the Greenville Healthy Neighborhoods Project, 2014.^a

	M (SD) or Percentage
Body mass index	30.3 (7.2)
Neighborhood social capital
Social cohesion (range: 1-5)	3.3 (0.7)
Collective efficacy (range: 1-5)	3.1 (1.1)
Support from neighbors (range: (1-5)	3.5 (0.8)
Network social capital
Network diversity (range: 0-12)	4.1 (3.4)
Educational attainment
Less than high school	17.5
High school diploma	43.3
Some college	26.7
Bachelors’ degree or higher	12.5
Female	70.8
Age	56.0 (14.6)
Marital status
Married/cohabiting	23.1
Single/separated/divorced/widowed	76.9
Employment status
Employed	32.7
Unemployed/disabled/retired	67.3
Household income
Less than $15 000	47.2
$15 000-$29 999	22.5
$30 000-$44 999	6.5
$45 000-$59 999	14.3
$60 000+	9.5

Abbreviation: SD, standard deviation.

^a n = 337.

Measures

Dependent variable

Body mass index was calculated using self-reported height and weight. Participants reported their height in feet and inches and their weight in pounds which was converted to a BMI score using the standard equation for adults: BMI = [weight (lbs.) / height (in.)²] × 703 kg/m² (in²/lbs). The distribution was fairly normal, so a raw BMI score was used in all analyses. In a previous cohort study, self-reported height and weight data that were used to calculate BMI scores were shown to be valid measures for examining relationships in epidemiological studies.⁴⁹

Neighborhood social capital

As is described further, neighborhood social capital was assessed using three comprehensive scales employed most commonly to measure neighborhood social environments: social cohesion, collective efficacy, and social support from neighbors. Traditionally, neighborhood capital is assessed at the aggregate or community level and is then often applied to individual health outcomes via multilevel models.²² However, in this exploratory study of social capital among black residents of the Southeastern United States, neighborhood-level data were not available. Following prior studies that have also lacked aggregated data,^50,51 neighborhood capital was measured at the individual level and from within the same sample in which health outcomes were reported. Correlation between the three items was low (social cohesion and collective efficacy: r = .3; social cohesion and social support: r = .4; and collective efficacy and social support: r = .2), suggesting that each scale captured a unique construct comprising different elements of neighborhood capital.

Social cohesion was measured using a five-item scale that assessed perceived trust and shared values.¹² The scale included items such as “People in this neighborhood can be trusted” and “People in this neighborhood do not share the same values,” which were rated on a five-point scale (1 = strongly disagree, 5 = strongly agree). Two items were reverse-coded, and a mean score was calculated across the five items such that higher scores represented greater social cohesion. The scale has shown good reliability and validity in previous studies among similar populations.⁵² Cronbach α for these five items in the current study was α = 0.71, similar to previous studies that have utilized this scale.^53,54

Collective efficacy was measured using a five-item scale that assessed a participant’s perceptions about the willingness of their neighbors to intervene on behalf of the common good.¹² The scale included items such as “How likely would neighbors be to intervene if children were hanging out in the neighborhood or around a school at night?” Participants were asked to rate each of the scenarios using a five-point likeliness scale (1 = very unlikely, 5 = very likely). A mean score was calculated across the five items where higher scores represented greater collective efficacy. In the current study, Cronbach α across these five items (α = 0.87) was identical to previously reported data among Black women living in the US South.⁵⁵

Social support from neighbors was used as an additional measure of neighborhood social capital. Participants were asked to rate their level of agreement using a five-point Likert-type scale (1 = strongly disagree, 5 = strongly agree) for four items about different types of support they might receive from their neighbors, such as information and advice about job opportunities. These items, adapted from the Montreal Neighborhood Networks and Healthy Aging Panel Study,⁵⁶ were assessed for internal consistency, and a mean score was calculated across all four items. Similar to previous studies that assessed social support,^53,55 Cronbach α across these four items was α = 0.80.

Network social capital

Network social capital was assessed using a position generator. Position generators are a common survey tool used to measure individual-level social capital and to specifically capture access to social resources.⁵⁷ The position generator asks respondents to identify whether they are on a “first name basis” with people holding a range of occupations in society, such as an accountant, physician, or high school teacher. The twelve occupations in the position generator instrument have previously been assigned a prestige value⁵⁸ that serves as an indicator of accessible social capital.⁵⁷ For example, the occupation with the lowest prestige score was a machine operator (20.83), and the highest prestige score was a physician (86.05). From the position generator, occupation-based network diversity was calculated as the number of unique occupations accessed (range 0-12) and was used as a continuous measure of network capital. While other measures can be extracted from the position generator, including the highest prestige score accessible (ie, reach), prior studies using these measures indicate network diversity is most commonly associated with health outcomes.¹⁶ Table 2 provides detailed information about each of the twelve positions, their prestige scores, and the total percentage of participants who reported they had access to each position.

Table 2.

Network Diversity Characteristics by Educational Attainment of the Greenville Healthy Neighborhoods Project, 2014.^a

			Access by Educational Attainment (%)
Occupation	Prestige Score	Total Access, %	Less than HS, n = 57	HS Diploma/GED, n = 145	Some College, n = 89	Bachelors’ Degree or Higher, n = 42
Machine Operator	20.83	33.3	14.0	35.2	39.3	40.5
Janitor	22.33	38.7	26.3	40.0	42.7	42.9
Store cashier	29.45	44.4	26.3	43.5	59.6	40.5
Carpenter	38.92	36.9	24.6	34.5	42.7	50.0
Receptionist	39.02	29.7	10.5	29.0	41.6	33.3
Mechanic	39.64	50.5	35.1	49.7	58.4	57.1
Welder	41.89	12.6	10.5	11.0	20.2	4.8
Musician/artist	46.56	38.4	21.1	30.3	46.1	73.8
Accountant	65.38	18.0	10.5	14.5	24.7	26.2
Registered nurse	66.48	39.6	19.3	35.9	50.6	57.1
High school teacher	73.51	42.3	19.3	39.0	50.6	71.4
Physician	86.05	21.6	10.5	17.2	27.0	48.5

^a n = 337.

Participant characteristics

Demographic characteristics provided as part of the survey included age (continuous), gender (male or female), employment status (employed or unemployed/disabled/retired), and marital status (married/cohabiting or single/separated/divorced/widowed). Household income was reported as a categorical variable and was collapsed into four categories: very low (less than $15 000), low ($15 000-$29 999), middle ($30 000-$59 999), and high ($60 000 or higher). Educational attainment was reported as the highest level of education completed and was collapsed into the following four categories: less than high school, high school diploma or GED, some college or associate’s (2-year) degree, and college (4-year) or graduate degree.

Analysis

Descriptive statistics were used to explore demographic characteristics of the sample as well as the distribution of key study variables. In answering our primary study questions, multilevel multivariable regression models were used to first examine the relationship between socioeconomic status (SES; income and education) and each of the measures of neighborhood and network social capital. Linear regression models were performed for each of the three neighborhood social capital outcomes, whereas Poisson regression was used to handle network diversity (count variable; range 0-12). Both income and education were entered into the models simultaneously. Each model adjusted for age (continuous), gender, marital status, and employment status. Model estimations were performed using the mixed (linear) and meqrpoisson (count) commands in STATA software version 14.2.

Next, multilevel linear regression models were used to examine the relationship between social capital and BMI. All three measures of neighborhood social capital were entered into model 1. Network social capital was independently tested in model 2, and both forms of capital were entered into model 3. Each model also adjusted for age, gender, marital status, and employment status. All model estimations were performed using the mixed command in STATA software version 14.2.

Multilevel modeling was used to account for the hierarchical nature of RDS data and nonindependence of the sample.⁵⁹ First, a 3-level model of individuals with sampling chains within neighborhoods was assessed. Because no variance was found at the neighborhood cluster level, the data were estimated utilizing a 2-level model (individuals within sampling chains). Additionally, a robust (sandwich) covariance estimator was used to account for unknown clustering of observations within neighborhoods and sampling chains. This method uses a “working covariance matrix” during the estimation step and the sandwich covariance estimator to construct confidence intervals and P values. Sandwich estimators are useful under circumstances when the correlation structure among observations, such as participants within an RDS-based sample, is unknown.⁶⁰

Missing data were handled using STATA’s multiple imputation command (mi impute) with the chained equations option.⁶¹ Observations with missing data on income (n = 35), age (n = 1), gender (n = 1), education (n = 3), and social capital measures (eg, network diversity n = 9) were imputed a total of 50 times based on chained equations and predictors with complete data (eg, BMI score and marital status). The STATA command mi estimate was used to replicate the subsequent regression analyses across the 50 computed values.

Results

Sample Characteristics

Similar to previous studies using RDS,⁶² the current study had a 1:1 recruitment ratio, meaning that, on average, each participant recruited 1 additional participant. In total, 430 residents completed the survey across four waves of recruitment in the eight neighborhoods. A total of 93 observations were dropped due to missing data (BMI scores: n = 34) or reporting a race other than black (n = 59) for a final sample size of 337.

Descriptive statistics are presented in Table 1. Participants were, on average, 56 years old (SD = 14.6). More than two-thirds of the sample was female (70.8%), over half had a high school education (43.3%) or less (17.5%), and nearly half had an annual household income of less than $15 000 (47.2%).

Access to Social Capital

Table 2 provides detailed information about the twelve occupations that comprised the position generator scale. While half of the sample knew (ie, had access to) a mechanic (50.5%), less than a fifth of the sample knew an accountant (18.0%) on a first-name basis. The percentage of participants who knew each occupation is also listed by participant education level. In general, participants with higher levels of education (ie, a college degree) were more likely to know someone from each of the listed occupations than those with less than a high school diploma.

To examine access to social capital by SES, Table 3 presents results from the multilevel model regression analyses examining the relationship between educational attainment and household income with neighborhood and network social capital measures, while controlling for demographic characteristics. Compared to participants with less than a high school diploma, higher levels of education at any level were associated with greater network diversity (high school: incident rate ratio [IRR] = 1.79, 95% confidence interval [CI]: 1.37-2.33; some college: IRR = 2.26, 95% CI: 1.71-2.98; college/graduate degree: IRR = 2.15, 95% CI: 1.59-2.90). Household income was not associated with network social capital, but was associated with each measure of neighborhood social capital. Compared to those with lowest household income, participants with the highest income level (≥$60 000) reported greater social cohesion among their neighbors (b = 0.43, 95% CI: 0.10-0.76). This association was also found for collective efficacy (b = 0.73, 95% CI: 0.19-1.26) and for social support from neighbors (b = 0.39, 95% CI: 0.07-0.71).

Table 3.

Multilevel Linear (Neighborhood Social Capital) and Poisson (Network Social Capital) Estimates of Associations Between Socioeconomic Status Indicators and Social Capital.^a

	Neighborhood Social Capital, b (95% CI)			Network Social Capital, IRR (95% CI)
	Social Cohesion	Collective Efficacy	Support from Neighbors	Network Diversity
Educational attainment
Less than high school	−	−	−	−
High school diploma	−0.00 (−0.21 to 0.20)	0.24 (−0.10 to 0.58)	0.18 (−0.11 to 0.47)	1.79 (1.37 to 2.33)^b
Some college	0.03 (−0.20 to 0.26)	0.07 (−0.25 to 0.40)	−0.06 (−0.32 to 0.20)	2.26 (1.71 to 2.98)^b
Bachelor’s degree or higher	−0.07 (−0.37 to 0.23)	0.05 (−0.42 to 0.52)	−0.01 (−0.30 to 0.27)	2.15 (1.59 to 2.90)^b
Household income
Less than $15 000	−	−	−	−
$15 000−$29 999	0.24 (0.03 to 0.44)^c	0.35 (0.06 to 0.65)^c	0.02 (−0.29 to 0.33)	1.05 (0.84 to 1.32)
$30 000−$44,999	0.26 (−0.10 to 0.63)	−0.07 (−0.57 to 0.44)	−0.22 (−0.64 to 0.20)	1.22 (0.93 to 1.61)
$45 000−$59,999	0.30 (0.04 to 0.55)^c	0.60 (0.18 to 1.02)^b	0.19 (−0.17 to 0.56)	1.22 (0.97 to 1.54)
$60 000+	0.43 (0.10 to 0.76)^b	0.73 (0.19 to 1.26)^b	0.39 (0.07 to 0.71)^b	1.17 (0.87 to 1.56)
Intercept	2.71 (2.32 to 3.11)	2.63 (2.07 to 3.20)	3.15 (2.60 to 3.70)	−

Abbreviation: CI, confidence interval; IRR, incident rate ratio.

^a n = 337. Models control for age, gender, marital status, and employment status. Data: Greenville Healthy Neighborhoods Project, 2014.

^b P < .01.

^c P < .05.

Social Capital and BMI

Table 4 presents results from the multilevel linear regression models examining the relationship between neighborhood social capital and network social capital and BMI. Model 1, which examines measures of neighborhood social capital, indicates that higher social cohesion is associated with lower BMI (b = −1.26, 95% CI: −2.39 to −0.13) after controlling for sociodemographic characteristics. Neither collective efficacy nor support from neighbors was associated with BMI. Model 2 indicates that greater network diversity is associated with higher BMI (b = 0.33, 95% CI: 0.09-0.56). Model 3 tests the two forms of capital simultaneously, and both of these associations remain statistically significant (neighborhood social capital: b = −1.25, 95% CI: −2.39 to −0.11; network social capital: b = 0.31, 95% CI: 0.08-0.55).

Table 4.

Multilevel Linear Estimates of Associations Between Neighborhood and Network Social Capital, and BMI.^a

	Model 1	Model 2	Model 3
	b (95% CI)	b (95% CI)	b (95% CI)
Neighborhood social capital
Social cohesion	−1.26 (−2.39 to −0.13)^b		−1.25 (−2.39 to −0.11)^b
Collective efficacy	0.32 (−0.27 to 0.92)		0.31 (−0.28 to 0.91)
Support from neighbors	0.83 (−0.06 to 1.72)		0.61 (−0.30 to 1.52)
Network social capital
Network diversity		0.33 (0.09 to 0.56)^c	0.31 (0.08 to 0.55)^c
Intercept	24.82 (18.91 to 30.73)	24.78 (20.40 to 29.15)	25.13 (19.34 to 30.92)

Abbreviation: CI, confidence interval. ^an = 337. Models control for age, gender, marital status, employment status, educational attainment, and annual household income. Data: Greenville Healthy Neighborhoods Project, 2014.

^b P < .05.

^c P < .01.

Discussion

The current study explores correlates of two forms of social capital among Black residents of economically disadvantaged neighborhoods in the US South. We found that higher household income was associated with greater neighborhood social capital, while higher levels of educational attainment were associated with greater network diversity. Although social cohesion appeared to be protective against overweight and obesity, greater network diversity was associated with higher BMI, net of income, and education. These findings provide several important contributions to the growing literature on the role of social capital for health.

First, the current study distinguishes between two forms of social capital – neighborhood and network – and provides evidence to suggest there may be differences in access to social capital based on an individual’s socioeconomic position. In support of previous hypotheses,^41,63 the current data indicate that individuals with higher educational attainment are more likely to have connections with persons in a wider range of occupations, including across both higher and lower occupational prestige scores. More generally, these data support the notion that individuals with higher SES, measured as either income or education, may have greater access to social capital and thus more leverage for improving health.^19,64 Additionally, our study underscores findings from previous research to indicate neighborhood capital is largely associated with household income,^6,14 whereas educational attainment may be more closely linked with social network characteristics,^65,66 including network diversity. The distinction between these two pathways may be critical for understanding the mixed role of social capital on health, described below.

Second, the findings suggest neighborhood and network social capital may work in opposite ways in association with BMI. Social cohesion, a neighborhood form of social capital, was associated with lower BMI. This is consistent with prior studies that find neighborhood social capital is protective of health and associated with better health outcomes, including self-rated health,⁵ health behaviors,⁶⁷ and mental health.⁴ The current findings also corroborate prior research to indicate neighborhood capital is associated with lower obesity. While a recent literature review found dozens of studies conducted primarily among white and Hispanic populations,²⁸ the current findings suggest neighborhood capital may benefit Black Americans as well.

However, unlike a previous study which found a protective relationship between network social capital and BMI,⁷ the current study found that greater network capital was associated with higher BMI. Of note, the prior study was conducted among residents of Montreal where the overall prevalence of obesity was 21.5% in 2012,⁶⁸ whereas the prevalence of obesity among the current sample of Black residents of Greenville was 48.6%. This large discrepancy in prevalence among these two groups may have important implications for the observed differences in relationships between network capital and BMI. For example, social and cultural norms associated with obesity likely differ between residents of Montreal versus residents of Greenville and more broadly, the Southeastern United States. If network diversity is associated with BMI via social norms, then network diversity is likely less protective of overweight and obesity among populations where obesity is more common. For example, network diversity has also been associated with higher odds of binge drinking,²⁹ frequently a social behavior, highlighting the detrimental impact social capital may have on health through increased social pressure or conformity to social norms. Future studies should explore possible mechanisms, including social influence and control, as well as outcomes most susceptible to peer influence, to shed additional light on the potentially negative relationship between social capital and health.

The positive association between network diversity and BMI in this study may also underscore the uniqueness of these relationships among black and low-income populations, in particular. Limited research has examined the utility of social capital for health among blacks and low-income populations, despite work to indicate that levels of social capital differ among these groups as compared to whites and more affluent populations,⁴¹ limiting our understanding of whether or how social capital is related to health among poor and black populations in the US South. Furthermore, there may be other types of social capital more relevant for health among black populations. For example, a substantial body of work suggests neighborhood ethnic heterogeneity, often an indicator of higher socioeconomic status among black residents,⁶⁹ is associated with civic engagement and generalized trust,^70,71 and may be beneficial for black residents particularly.⁷² These two measures of social capital have been studied in association with health^73,74 but not specifically among black populations. As such, future studies may seek to explore these potential mechanisms, including measures of generalized trust, in association with chronic disease and BMI among black and low-income populations living in the US South.

Limitations

The results from this exploratory study should be interpreted in light of certain limitations. In addition to the inability to examine aggregate levels of neighborhood social capital discussed previously, the cross-sectional nature of the data does not allow us to determine the directionality of the findings, such as whether education and income contribute to social capital, or vice versa. Similarly, it is unknown whether social capital affects BMI or whether overweight or obesity influence social capital in the current study. This is an ongoing conversation within the literature, since it has been shown that poor health is associated with lower social engagement.^75,76 Further, levels of social capital likely change over time and throughout the life course to impact health outcomes, and as such, cross-sectional measures of capital may limit the ability to detect important relationships. Second, this study was limited by a relatively small sample size, which may also hamper the ability to detect statistically significant associations or generalize the findings for broader application. However, this sample represents an important and understudied population within the literature and the findings here warrant further investigation with similar samples. Finally, the nature of the sampling methodology may have selected individuals who were more socially integrated, which may also limit the generalizability of the findings. The RDS methodology requires that participants be invited into the study by another community member, which would likely not capture individuals who are socially isolated. However, RDS is an ideal sampling strategy for engaging populations who may have been reluctant to participate in research otherwise.⁵⁹ As such, the data represent an opportunity to explore potential relationships between social capital and BMI among a sample of black Americans with inequitably high rates of obesity.

Conclusions

In contrast to a body of work that suggests single indicators of social capital may be sufficient for measurement in the context of health, the current data reiterate there are distinct forms of social capital which are differentially associated with health. Results from the current analyses reveal inequity in access to social capital by household income and educational attainment. By examining two distinct forms of social capital, namely neighborhood and network measures, our study suggests each form of social capital may be differentially associated with BMI among black adults living in the Southeastern United States. This has important ramifications for thinking about how social relationships can be utilized in response to chronic health conditions, including obesity. Current policies and future interventions aimed at increasing social capital to improve population health may be better informed by a deeper, richer understanding of the various forms of social capital, how individuals gain access to these forms of capital, and their relationships with obesity, particularly among diverse and underrepresented population groups.

So What?

What is already known on this topic?

Social capital is important for health, including chronic health outcomes such as obesity, although some research indicates the links between social capital and health are not always beneficial. Most studies only examine a single dimension of social capital. Further, access to social capital is hypothesized to vary based on socioeconomic status.

What does this article add?

The current study examines associations between socioeconomic status and two forms of social capital, namely, neighborhood and network measures as well as how these distinct forms of social capital are associated with body mass index (BMI) among black residents of low-income communities. The findings indicate those with less education and less income have lower levels of network and neighborhood capital, respectively. Further, while neighborhood capital is associated with lower BMI, network capital is associated with higher BMI among black Americans living in resource-poor communities.

What are the implications for health promotion practice or research?

This study has important ramifications for how social relationships and social capital can be utilized in response to chronic health conditions, including obesity, and particularly among black Americans. The assessment of multiple dimensions of social capital among diverse populations is needed to better inform health promotion research on the link between social capital and health. The findings suggest neighborhood social capital may be a fruitful target for health promotion practices and policies seeking to leverage social relationships and environments for the improvement in chronic health conditions among black Americans

Footnotes

Acknowledgments

The authors wish to thank the leaders of the eight participating neighborhoods for their contributions to the Greenville Healthy Neighborhoods Project.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Blue Cross Blue Shield Foundation of South Carolina and the University of South Carolina Office of the Vice President for Research.

ORCID iD

Stephanie T. Child

References

Moore

Kawachi

. Twenty years of social capital and health research: a glossary. J Epidemiol Community Health. 2017;71(5):513–517. doi:10.1136/jech-2016-208313

Carpiano

RM.

Toward a neighborhood resource-based theory of social capital for health: can Bourdieu and sociology help?

Soc Sci Med. 2006;62(1):165–175. doi: 10.1016/j.socscimed.2005.05.020

Nyqvist

Pape

Pellfolk

Forsman

Wahlbeck

. Structural and cognitive aspects of social capital and all-cause mortality: a meta-analysis of cohort studies. Soc Indic Res. 2013;116(2):545–566. doi: 10.1007/s11205-013-0288-9

Riumallo-Herl

Kawachi

Avendano

. Social capital, mental health and biomarkers in chile: assessing the effects of social capital in a middle-income country. Soc Sci Med. 2014;105:47–58. doi: 10.1016/j.socscimed.2013.12.018

Giordano

Björk

Lindström

. Social capital and self-rated health—a study of temporal (causal) relationships. Soc Sci Med. 2012;75(2):340–348. doi: 10.1016/j.socscimed.2012.03.011

Mackenbach

Lakerveld

van Oostveen

, et al. The mediating role of social capital in the association between neighbourhood income inequality and body mass index. Eur J Public Health. 2017;27(2):218–223. doi: 10.1093/eurpub/ckw157

Moore

Daniel

Paquet

Dubé

Gauvin

. Association of individual network social capital with abdominal adiposity, overweight and obesity. J Public Health. 2009;31(1):175–183. doi: 10.1093/pubmed/fdn104

Suglia

Shelton

Hsiao

Wang

Rundle

Link

. Why the neighborhood social environment is critical in obesity prevention. J Urban Health. 2016;93(1):206–212. doi: 10.1007/s11524-015-0017-6

Holtgrave

Crosby

. Is social capital a protective factor against obesity and diabetes? findings from an exploratory study. Ann Epidemiol. 2006;16(5):406–408. doi: 10.1016/j.annepidem.2005.04.017

10.

Veitch

van Stralen

Chinapaw

, et al. The neighborhood social environment and body mass index among youth: a mediation analysis. Int J Behav Nutr Phys Act. 2012;9:31. doi: 10.1186/1479-5868-9-31

11.

Putnam

Leonardi

Nanetti

. Making democracy work: civic traditions in modern Italy. Princeton, NJ: Princeton University Press; 1994.

12.

Sampson

Raudenbush

Earls

. Neighborhoods and violent crime: a multilevel study of collective efficacy. Science. 1997;277(5328):918–924. doi: 10.1126/science.277.5328.918

13.

Sampson

Morenoff

Earls

. Beyond social capital: spatial dynamics of collective efficacy for children. Am Sociol Rev. 1999;64(5):633–660. doi: 10.2307/2657367

14.

Kawachi

Kennedy

Lochner

Prothrow-Stith

. Social capital, income inequality, and mortality. Am J Public Health. 1997;87(9):1491–1498. doi: 10.2105/AJPH.87.9.1491

15.

Moore

Haines

Hawe

Shiell

. Lost in translation: a genealogy of the “social capital” concept in public health. J Epidemiol Community Health. 2006;60(8):729–734. doi: 10.1136/jech.2005.041848

16.

Legh-Jones

Moore

. Network social capital, social participation, and physical inactivity in an urban adult population. Soc Sci Med. 2012;74(9):1362–1367. doi: 10.1016/j.socscimed.2012.01.005

17.

Mohnen

Völker

Flap

Subramanian

Groenewegen

. The influence of social capital on individual health: is it the neighbourhood or the network? Soc Indic Res. 2015;121(1):195–214. doi: 10.1007/s11205-014-0632-8

18.

Moore

Shiell

Hawe

Haines

. The privileging of communitarian ideas: citation practices and the translation of social capital into public health research. Am J Public Health. 2005;95(8):1330–1337. doi: 10.2105/AJPH.2004.046094

19.

Berkman

Glass

. Social integration, social networks, social support, and health. In: Berkman

Kawachi

, eds. Social Epidemiology. New York, NY: Oxford University Press; 2000:137–173.

20.

Lin

. Social networks and status attainment. Annu Rev Sociol. 1999;25(1):467–487. doi: 10.1146/annurev.soc.25.1.467

21.

Moore

Bockenholt

Daniel

Frohlich

Kestens

Richard

. Social capital and core network ties: a validation study of individual-level social capital measures and their association with extra- and intra-neighborhood ties, and self-rated health. Health Place. 2011;17(2):536–544. doi: 10.1016/j.healthplace.2010.12.010

22.

Cohen

Finch

Bower

Sastry

. Collective efficacy and obesity: the potential influence of social factors on health. Soc Sci Med. 2006;62(3):769–778. doi: 10.1016/j.socscimed.2005.06.033

23.

Leyden

. Social capital and the built environment: the importance of walkable neighborhoods. Am J Public Health. 2003;93(9):1546–1551. doi: 10.2105/AJPH.93.9.1546

24.

McNeill

Kreuter

Subramanian

. Social environment and physical activity: a review of concepts and evidence. Soc Sci Med. 2006;63(4):1011–1022. doi: 10.1016/j.socscimed.2006.03.012

25.

Sallis

Bowles

Bauman

, et al. Neighborhood environments and physical activity among adults in 11 countries. Am J Prev Med. 2009;36(6):484–490. doi: 10.1016/j.amepre.2009.01.031

26.

De Jesus

Puleo

Shelton

Emmons

. Associations between perceived social environment and neighborhood safety: health implications. Health Place. 2010;16(5):1007–1013. doi: 10.1016/j.healthplace.2010.06.005

27.

Matthews

Yang

T-C

. Exploring the role of the built and social neighborhood environment in moderating stress and health. Ann Behav Med Publ Soc Behav Med. 2010;39(2):170–183. doi: 10.1007/s12160-010-9175-7

28.

Carrillo-Álvarez

Kawachi

Riera-Romaní

. Neighbourhood social capital and obesity: a systematic review of the literature. Obes Rev. 2019;20(1):119–141. doi: 10.1111/obr.12760

29.

Child

Stewart

Moore

. Perceived control moderates the relationship between social capital and binge drinking: longitudinal findings from the montreal neighborhood networks and health aging (MoNNET-HA) panel. Ann Epidemiol. 2017;27(2):128–134. doi: 10.1016/j.annepidem.2016.11.010

30.

Chuang

. Gender differences in relationships between social capital and individual smoking and drinking behavior in Taiwan. Soc Sci Med. 2008;67(8):1321–1330. doi: 10.1016/j.socscimed.2008.06.033

31.

Pförtner

T-K

Clercq

Lenzi

, et al. Does the association between different dimension of social capital and adolescent smoking vary by socioeconomic status? a pooled cross-national analysis. Int J Public Health. 2015;60(8):901–910. doi: 10.1007/s00038-015-0734-3

32.

De Clercq

Pfoertner

T-K

Elgar

Hublet

Maes

. Social capital and adolescent smoking in schools and communities: a cross-classified multilevel analysis. Soc Sci Med 1982. 2014;119:81–87. doi: 10.1016/j.socscimed.2014.08.018

33.

Domínguez

Watkins

. Creating networks for survival and mobility: social capital among African-American and Latin-American low-income mothers. Soc Probl. 2003;50(1):111–135. doi: 10.1525/sp.2003.50.1.111

34.

Brown

. Faith and works: church-based social capital resources and African American political activisim. Soc Forces. 2003;82(2):617–641. doi: 10.1353/sof.2004.0005

35.

Lochner

Kawachi

Brennan

Buka

. Social capital and neighborhood mortality rates in Chicago. Soc Sci Med. 2003;56(8):1797–1805. doi: 10.1016/S0277-9536(02)00177-6

36.

Hummer

Chinn

. RACE/ETHNICITY AND U.S. ADULT MORTALITY. Bois Rev Soc Sci Res Race. 2011;8(1):5–24. doi: 10.1017/S1742058X11000051

37.

Piccolo

Yang

Bliwise

Yaggi

Araujo

. Racial and socioeconomic disparities in sleep and chronic disease: results of a longitudinal investigation. Ethn Dis. 2013;23(4):499–507.

38.

Olshansky

Antonucci

Berkman

, et al. Differences In life expectancy due to race and educational differences are widening, and many may not catch up. Health Aff (Millwood). 2012;31(8):1803–1813. doi: 10.1377/hlthaff.2011.0746

39.

Centers for Disease Control and Prevention. Summary health statistics for U.S. adults: national health interview survey; 2010. accessed from 2012. http://www.cdc.gov/nchs/data/series/sr_10/sr10_252.pdf.

40.

Roger

Fullerton

Lloyd-Jones

, et al. Executive summary: heart disease and stroke statistics—2012 update a report from the American heart association. Circulation. 2012;125(1):188–197. doi: 10.1161/CIR.0b013e3182456d46

41.

Lin

. Inequality in social capital. Contemp Sociol. 2000;29(6):785–795. doi: 10.2307/2654086

42.

Lin

. Social capital: a theory of social structure and action. Cambridge, England: Cambridge University Press; 2002.

43.

Portes

. Social capital: its origins and applications in modern sociology. In: Lesser

, ed. Knowledge and Social Capital: Foundations and Applications. Amsterdam, the Netherlands: Elsevier; 2000:1–24.

44.

Ioannides

Loury

. Job information networks, neighborhood effects, and inequality. J Econ Lit. 2004;42(4):1056–1093.

45.

McPherson

Smith-Lovin

Cook

. Birds of a feather: homophily in social networks. Annu Rev Sociol. 2001;27:415–444.

46.

Malekinejad

Johnston

Kendall

Kerr

LRFS

Rifkin

Rutherford

. Using respondent-driven sampling methodology for hiv biological and behavioral surveillance in international settings: a systematic review. AIDS Behav. 2008;12(1):105–130. doi: 10.1007/s10461-008-9421 -1

47.

Schonlau

Liebau

. Respondent-driven sampling. Stata J. 2012;12(1):72–93.

48.

U. S. Census Bureau. 2010-2014 American community survey 5-year estimates; 2014. http://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?fpt=table. Accessed July 15, 2016.

49.

Spencer

Appleby

Davey

Key

. Validity of self-reported height and weight in 4808 EPIC–Oxford participants. Public Health Nutr. 2002;5(04):561–565. doi: 10.1079/PHN2001322

50.

Nesbit

Kolobe

Sisson

Ghement

. A model of environmental correlates of adolescent obesity in the United States. J Adolesc Health. 2014;55(3):394–401. doi: 10.1016/j.jadohealth.2014.02.022

51.

Poortinga

. Perceptions of the environment, physical activity, and obesity. Soc Sci Med. 2006;63(11):2835–2846. doi: 10.1016/j.socscimed.2006.07.018

52.

Raudenbush

Sampson

. Ecometrics: toward a science of assessing ecological settings, with application to the systematic social observation of neighborhoods. Sociol Methodol. 1999;29(1):1–41. doi: 10.1111/0081-1750.00059

53.

Alegria

Sribney

Mulvaney-Day

. Social cohesion, social support and health among Latinos in the United States. Soc Sci Med 1982. 2007;64(2):477–495. doi: 10.1016/j.socscimed.2006.08.030

54.

Mujahid

Roux

AVD

Morenoff

Raghunathan

. Assessing the measurement properties of neighborhood scales: from psychometrics to ecometrics. Am J Epidemiol. 2007;165(8):858–867. doi: 10.1093/aje/kwm040

55.

Andersen

Gustat

Becker

. The relationship between the social environment and lifestyle-related physical activity in a low-income African American inner-city southern neighborhood. J Community Health. 2015;40(5):967–974. doi: 10.1007/s10900-015-0019-z

56.

Moore

Buckeridge

Dubé

. Cohort profile: the montreal neighbourhood networks and healthy aging (MoNNET-HA) study. Int J Epidemiol. 2014;45(1):45–53. doi: 10.1093/ije/dyu137

57.

Van der Gaag

Snijders

Flap

. Position generator measures and their relationship to other social capital measures. Soc Cap Int Res Program. 2008:27–49.

58.

Van der Gaag

Snijders

Flap

. Position generator measures and their relationship to other social capital measures. In: Lin

Erickson

, eds. Social Capital: An International Research Program. Oxford; 2008:27–49.

59.

Rhodes

McCoy

. Condom use among immigrant Latino sexual minorities: multilevel analysis after respondent-driven sampling. AIDS Educ Prev. 2015;27(1):27–43. doi: 10.1521/aeap.2015.27.1.27

60.

Kauermann

Carroll

. The sandwich variance estimator: efficiency properties and coverage probability of confidence intervals; 2000. https://epub.ub.uni-muenchen.de/1579/. Accessed November 13, 2015.

61.

White

Royston

Wood

. Multiple imputation using chained equations: issues and guidance for practice. Stat Med. 2011;30(4):377–399. doi: 10.1002/sim.4067

62.

Frost

SDW

Brouwer

Cruz

MAF

, et al. Respondent-driven sampling of injection drug users in two U.S.–Mexico border cities: recruitment dynamics and impact on estimates of HIV and syphilis prevalence. J Urban Health. 2006;83(1):83–97. doi: 10.1007/s11524-006-9104-z

63.

Gorman

Sivaganesan

. The role of social support and integration for understanding socioeconomic disparities in self-rated health and hypertension. Soc Sci Med. 2007;65(5):958–975. doi: 10.1016/j.socscimed.2007.04.017

64.

Uphoff

Pickett

Cabieses

Small

Wright

. A systematic review of the relationships between social capital and socioeconomic inequalities in health: a contribution to understanding the psychosocial pathway of health inequalities. Int J Equity Health. 2013;12:54. doi: 10.1186/1475-9276-12-54

65.

Fischer

. To dwell among friends: personal networks in town and city. Chicago, IL: University of Chicago Press; 1982.

66.

Fischer

Beresford

. Changes in support networks in late middle age: the extension of gender and educational differences. J Gerontol Ser B. 2015;70(1):123–131. doi: 10.1093/geronb/gbu057

67.

Nieminen

Prättälä

Martelin

, et al. Social capital, health behaviours and health: a population-based associational study. BMC Public Health. 2013;13:613. doi: 10.1186/1471-2458-13-613

68.

Navaneelan

Janz

. Adjusting the scales: obesity in the canadian population after correcting for respondent bias. Statistics Canada; 2014. http://www.statcan.gc.ca/pub/82-624-x/2014001/article/11922-eng.htm. Accessed July 19, 2017.

69.

Clark

WAV

. Race, class, and place: evaluating mobility outcomes for African Americans. Urban Aff Rev. 2007;42(3):295–314. doi: 10.1177/1078087406292531

70.

Stolle

Soroka

Johnston

. When does diversity erode trust? neighborhood diversity, interpersonal trust and the mediating effect of social interactions. Polit Stud. 2008;56(1):57–75. doi: 10.1111/j.1467-9248.2007.00717.x

71.

Putnam

. E pluribus unum: diversity and community in the twenty-first century the 2006 Johan Skytte prize lecture. Scand Polit Stud. 2007;30(2):137–174. doi: 10.1111/j.1467-9477.2007.00176.x

72.

Marschall

Stolle

. Race and the city: neighborhood context and the development of generalized trust. Polit Behav. 2004;26(2):125–153. doi: 10.1023/B: POBE.0000035960.73204.64

73.

Hamamura

LMW

Chan

. The association between generalized trust and physical and psychological health across societies. Soc Indic Res. 2016;134:277–286. doi: 10.1007/s11205-016-1428-9

74.

Campos-Matos

Subramanian

Kawachi

. The ‘dark side’ of social capital: trust and self-rated health in European countries. Eur J Public Health. 2016;26(1):90–95. doi: 10.1093/eurpub/ckv089

75.

Harwood

Pound

Ebrahim

. Determinants of social engagement in older men. Psychol Health Med. 2000;5(1):75–85. doi: 10.1080/135485000106025

76.

Rosso

Taylor

Tabb

Michael

. Mobility, disability, and social engagement in older adults. J Aging Health. 2013;25(4):617–637. doi: 10.1177/0898264313482489