Significance

Latinas, and especially Latina immigrants, experience physical and mental health disparities at high rates. Social capital can help to mitigate these disparities and is shown to improve health for women, Latinas, children, and immigrants. Findings from the current study suggest that social capital mediates the effects of age, education, and economic well-being on the health of Latinas. The presence of social capital predicted improved health for Latinas who are more vulnerable (younger, less educated, and poorer). Most notably, social capital may have more relevance to the health and well-being of foreign-born Latinas than those who are native-born.

Introduction

Latinas experience higher risk of some health disorders (e.g., cervical cancer and diabetes) than non-Latino white women [10]. They also encounter extensive barriers to accessing health services [27]. Latinas who are foreign-born may be at an even greater risk for health disparities, due to a reduced likelihood of being proficient in English, having health insurance, or possessing legal documentation status [26, 28]. Social capital may serve as a protective factor, as it predicts improved physical and mental health for women [2], children [25], Latinas [7], and immigrants [32]. Gaining a more in depth understanding of the relationship between social capital and health and well-being for the native- and foreign-born Latina population provides important information to better address their health needs.

Social Capital

Social capital can be defined as “the aggregate of the actual or potential resources” that result from social connections [6]. Participation in social networks yields increased social contacts and supports [12], as well as community-based connections [31]. Two key forms of social capital are thought to emerge from one’s social networks: bonding social capital and bridging social capital [31]. Bonding social capital is accumulated through one’s closest networks (e.g., friends), while bridging social capital is accumulated through more distant connections or acquaintances [31]. Though some scholars do not differentiate between these two forms of social capital, others have found particular benefits of bonding social capital for self-rated health and reduced stress when compared to bridging social capital [5]. Bonding social capital relies on strong ties with individuals whom one already knows and is particularly beneficial for provision of emotional support or provision of assistance in emergency situations [16]. Thus, social capital, and especially bonding social capital, has received substantial attention for its impact on health and well-being [e.g., 5, 35].

Relationship to Health and Well-Being

Though a relationship between social capital and health is well supported, limited data exist on this correlation for Latinas. Additionally, the existing literature with Latinas does not differentiate between bonding and bridging social capital. Findings from a study of 99 Latina immigrants indicate that social capital predicted improved physical and mental health for non-students, however the opposite relationship was found for students [32]. An examination of the demographics for each group indicated that the non-student sample was younger, had fewer years of education, and was less likely to be employed. This finding suggests that social capital might provide greater advantage to more vulnerable respondents (i.e., younger, less educated, and unemployed).

Maternal social capital also benefits children’s health. Little relevant research exists in the United States, but studies in Indonesia and Vietnam demonstrate that maternal social capital predicted improved height-for-age and weight-for-age outcomes, and decreased risk of health disorders, for children [33]. The impact may be greater for children in low-income households and with less educated mothers [25]. Though social capital may be particularly relevant for vulnerable populations, it is not equally accessible to them [6].

Relevance for Vulnerable Populations

Social networks typically consist of members who are in similar socioeconomic situations [6], so that low-income or less socially connected (i.e., high social status) individuals have limited ability to benefit from wealthier, more well-connected networks.This dynamic can impact women [1] and Latinas [21], who are likely to belong to less powerful social circles than men or non-Hispanic whites. For example, women often engage with networks that provide non-monetary benefits, demanding more in terms of time and labor exchange than men’s networks that more readily demand financial resources [1]. Women’s connection might fail to financially benefit them to the same degree as men, but rather offer greater non-financial assets such as emotional or logistical support when needs arise. In a similar manner, Latino immigrants more readily come to the United States to work in positions of unskilled labor, preventing them from gaining access to more powerful networks. Despite this disparate access, social capital remains beneficial for Latinas and immigrants [30]. In fact, social capital might even be more important for these populations, as excluded groups such as Latinas and immigrants tend to form tighter networks that yield greater social capital benefits [14].

Predictors of Social Capital

Individual characteristics, such as age, education, and economic well-being, can predict one’s amount of social capital. Social capital is found to generally accrue with age, though some forms (such as daily contact with others) may decrease in older populations [22]. Older adults are more likely to lose weaker-based ties (e.g., professional and more distant acquaintances) as they age, but maintain their close friendships [15]. In terms of education, research indicates that higher educational attainment predicts greater social capital [18]. Acquiring more education increases knowledge and power, which can correspond with greater participation on local committees and increased social engagements (e.g., dinner parties) that yield higher social capital [18]. Economic well-being and social capital are also positively correlated, though most research has examined social capital as a predictor variable. As individuals in better financial situations have greater access to others with similar high socioeconomic statuses, one would expect that greater economic well-being predicts greater social capital.

Summary

In the current study, we examine the effect of bonding forms of social capital on maternal and child well-being for native- and foreign-born Latinas. Because bonding forms of social capital are built through tighter connections, which are particularly important for vulnerable populations, we anticipate that the ability to rely on various forms of emergency support will positively impact maternal and child health. We additionally examine the direct effects of age, education, and economic well-being on social capital and the indirect effects of these three variables on maternal and child well-being outcomes, with social capital mediating the impact. We hypothesize that age, education, and economic well-being will all positively predict social capital. Yet, we anticipate that the relationship will look different for native- and foreign-born Latinas, as immigrants are generally older, less educated, and more financially vulnerable [4, 34].

Methods

Data and Sample

Data were drawn from the baseline wave (1998–2000) and the 9-year follow up Wave 5 (2007–2010) of the Fragile Families and Child Well-being Study (FFCWB). This study represents an ongoing effort by Princeton University and Columbia University to follow a cohort of approximately 5000 infants born between the years 1998 and 2000 in a random stratified sample of large U.S. cities. Initially designed to develop an understanding of the lives of primarily unmarried birthmothers, the dataset is comprised of data collected over five waves across a nine-year timeframe from mothers, fathers, and children. Data were collected primarily through in-person and telephone interviews and contain demographic and lifestyle information, behavior, relationships, and physical and mental health. For this study, data were merged and utilized from both the mother core study questionnaires at baseline and Wave 5 and the child interviews at Wave 5.

Measures

All variables drawn from the dataset and used in subsequent analyses were binary or ordered categorical indicators with missing data being categorized as any missing response. All binary indicators were recoded so that an increase in their value was representative of a positive, or “yes” response. The study sample included 1333 Latina respondents, who were categorized as native- or foreign-born based on self-report.

Exogenous Variables

Variables that were not influenced by other variables within the hypothesized structural model included age, education, and economic well-being.

Age

Age was reported as a continuous variable in years. Native-born Latinas were significantly younger than those who were foreign-born. Only 1 of the 1333 cases had missing data for this variable.

Education

The educational attainment of Latinas was collected at baseline and reported as ordered categorical, ranging from one to four; one indicated the Latina had less than high school education (56.2 %), two indicated a high school or equivalent education (20.4 %), three indicated some college or technical education (19.5 %), and four indicated a college or graduate education (3.9 %). Native-born respondents were significantly more educated than those who were foreign-born. Only 2 of 1333 cases had missing data for this variable.

Economic Well-Being

The economic well-being variable was created as a proxy for measuring perceived adequacy of income level. This variable is comprised of four binary indicators of ability to meet: food needs (93 %), rent and mortgage payments (84 %), pay gas, power, or electric bills (77 %), and medical needs for all household members (93 %). A new dichotomous variable was created, with 0 indicating that a respondent was not able to meet at least one of the four basic needs and a 1 indicating that a respondent could meet all of the basic needs. The majority of respondents (67 %) could meet all basic needs. There were no missing data for this variable.

Endogenous Variables

Variables subject to influence from other variables within the model included the latent construct Social Capital, Maternal Well-being, and Child Well-being (see Table 1 for a listing of all observed indicators of latent constructs). Each of these three endogenous variables was selected a priori and then confirmed using a confirmatory factor analytic approach.

Table 1 Exogenous and endogenous variables in the model
Full size table

We utilized questions that assessed respondents’ access to, rather than use of, individual-level social capital [36]. We additionally drew from [11] measure of social capital, which also used questions from the Fragile Families dataset. These questions are reflective of the model posed by Wellman and Wortley [37] identifying five dimensions of bonding social capital, including emotional aid, small services, large services, financial aid, and companionship [19]. Additionally, the social capital measures assess participant’s ability to rely on emergency support, which is reflective of bonding social capital. Maternal well-being included questions related to physical and mental health, in addition to life satisfaction and control. Child well-being was comprised of questions on physical and mental health and relationship with mother.

Data Analysis

Data were analyzed using Mplus7, a program designed for the analysis of latent variables [24]. Because of the mixture of ordered categorical and continuous indicators within the model, analyses were performed using delta parameterization and weighted least squares and mean variance estimation [8, 20, 24]. Analyses were conducted sequentially. First, a confirmatory factor analytic model was tested to determine the factorial structure of the data. Second, the researchers tested the hypothesized structural model.

Modification indices were requested to examine the extent to which the model could be refined. All coefficients are reported as standardized unless otherwise noted. Standardized coefficients were produced using the STDYX option in Mplus7 and thus can be interpreted as effects in standard deviation units. Mediating effects were tested using the VIA argument, which allows the investigation of total indirect, specific indirect, and total effects of the hypothesized structural model.

Model Estimation

After confirming the hypothesized factorial structure of the data using a confirmatory factor analytic approach, a structural model was estimated based on the hypothesized relationships among exogenous and endogenous variables separately for native-born and foreign-born Latinas. First, Latinas’ age, education, and economic well-being were all hypothesized to be positively associated with social capital. Second, social capital was hypothesized to be positively associated with maternal well-being and child well-being. Third, a reciprocal relationship between maternal well-being and child well-being was hypothesized. Finally, it was hypothesized that social capital would mediate the effects of the exogenous indicators on maternal well-being and child well-being. We elected to use SEM for this analysis because of its strengths, including allowing for confirmatory testing of hypotheses, management of measurement and residual errors, and the ability to assess both direct and indirect pathways between the exogenous and endogenous variables [9].

Results

After accounting for a 34.2 % attrition rate from the baseline wave to wave 5 and handling missing data using full information maximum likelihood (FIML), our sample size was 874 and included 334 foreign-born (38 %) and 540 native-born (62 %) Latinas. Results indicate the a priori structural model suggest the model fit the data very well: X 2(246) = 329.631, p < .05; CFI = .968; TFI = .965; RMSEA = .028; WRMR = 1.339. Results of the model are in Table 2.

Table 2 Results of estimated structural model by origin (N = 874)
Full size table

For native-born Latinas, age, education, and economic well-being were not significant predictors of social capital, nor was social capital a significant predictor of maternal or child well-being. However, social capital significantly mediated the effects of age (STDYX = −.12, S.E. = .00, p < .05), education (STDYX = .26, S.E. = .07, p < .001), and economic well-being (STDYX = .26, S.E. = .06, p < .001) on maternal well-being, but not child well-being. No other significant relationships among latent constructs were observed in the native-born group.

For foreign-born Latinas, age, education, and economic well-being significantly predicted social capital. As hypothesized, education was positively associated with social capital (STDYX = .12, S.E. = .13, p < .05), and economic well-being was positively associated with social capital (STDYX = .27, S.E. = .13, p < .001), suggesting that foreign-born Latinas who were able to meet basic economic needs had significantly greater social capital. Age was negatively associated with social capital (STDYX = −.13, S.E. = .00, p < .05). Thus, being younger, having a higher education, and an ability to meet basic economic needs predicted greater social capital. Social capital was found to be a significant predictor of maternal well-being (STDYX = 1.37, S.E. = .21, p < .001), but not of child well-being. Participants with higher levels of social capital also reported improved well-being. Social capital mediated the effects of age (STDYX = −.18, S.E. = .00, p < .05), education (STDYX = .17, S.E. = .10, p < .001), and economic well-being (STDYX = .37, S.E. = .09, p < .001) on maternal well-being. No mediating effects were found for the relationship between age, education, and economic well-being and child well-being.

Discussion

This study contributes to our knowledge of social capital, health, and well-being for native- and foreign-born Latinas. Results suggest that, regardless of origin, social capital mediates the effects of age, education, and economic well-being on maternal health. Moreover, younger and foreign-born Latinas in this study who report higher educational attainment and an ability to meet their basic economic needs will have greater social capital, and thus report better overall health. Finally, for foreign-born Latinas, social capital is a strong predictor of maternal health, which suggests a relationship between overall health and social capital for this subgroup. In sum, foreign-born Latinas who are younger and are more educated are likely to have greater social capital, which in turn is attributable to improved overall health.

Predictors of Social Capital

Though not hypothesized, both age and economic well-being were negatively correlated with social capital for foreign-born Latinas. We suspect that each of these relationships relates to the process of immigration and settlement. Upon arrival in the United States, immigrants rely heavily on social networks (or prior immigrants) to secure housing, employment, and other necessities [13]. With increased time in the United States, and hence older age, comes improved economic stability in addition to a reduced reliance on social networks to meet basic well-being needs. The relationship between age and economic well-being and social capital was not present for native-born Latinas. This likely reflects the increased importance of social capital for more vulnerable populations, such as immigrants, who have reduced access to other resources (e.g., human and financial capital) [7]. The positive relationship between education and social capital for foreign-born Latinas was expected and supports findings from previous research [18]. Further research is needed to better understand predictors of social capital for native-born Latinas.

Social Capital and Well-Being

Consistent with prior studies, social capital predicted maternal well-being for foreign-born Latinas. This finding aligns with existing literature that emphasizes the relevance of social capital for immigrants [7]. Relying on community members to learn of affordable services or to obtain medical advice could be a necessary step for immigrants, who more likely lack health insurance or economic stability, to maintain health and well-being [21, 27]. Less expected was the lack of this relationship for native-born Latinas. Further investigation is needed to shed light on this outcome.

Social capital also served as a mediating variable between age, education, and economic well-being and maternal well-being for both immigrant and non-immigrant Latinas. This finding suggests that while research demonstrates that more vulnerable individuals (younger, less educated, and poorer) typically experience poorer health and well-being [17, 23], social capital might help to mediate this relationship. Thus, social capital may serve as an important resource for individuals who are vulnerable to increased risk of health concerns. Unexpectedly, no significant correlations existed between maternal social capital and child health.

Implications for Practice

Findings from this study highlight the importance of social capital for Latinas, especially those who are foreign-born, and have implications for service delivery. Health and social service programs often encounter limited resources to serve their consumers. Social capital is a low-cost resource that entities can promote as a prospective route to improved well-being. In fact, research demonstrates that programs and policies designed to increase access to social capital support improved health and well-being [29]. Both government and non-government programs can direct services at promoting social capital accumulation as a low-cost, but impactful resource for Latinas.

Building on previous research, specific forms of social capital that improve health and well-being include not only participation in groups, but also trust among members within these groups [3]. Use of group-centered services that bring Latinas together and encourage their reliance on one another, such as education, support, or treatment groups, could serve to facilitate beneficial measures of social capital. Additionally, linking consumers with community-based services (i.e., job training or English classes) also facilitates connection with larger networks as a means of improving access to social capital. Direct efforts at increasing connections among Latinas, and especially those who are foreign-born, may serve as a protective factor for their well-being.

A few study limitations should be noted. While the cross-sectional analysis provides important data, a longitudinal study would give temporal context to the relationship between social capital and well-being for Latinas and help to consider the effect of time spent in the United States on social capital accumulation. Another limitation is that the child health and well-being measures are limiting in that they do not give detailed data regarding the overall health of participants. Additionally, we did not control for acculturation or citizenship status, which are correlates of health outcomes, and were not able to examine immigrants by specific country of origin to evaluate distinctions in the role of social capital for Latinas originating from different locales. Future research should to examine the role of specific forms of social capital on health and well-being, while controlling for acculturation and citizenship status, in order to provide a more in depth understanding of the mechanisms through which well-being can be improved for this population.

Conclusion

This study is the first of its kind and sheds light on the importance of social capital for Latinas, and especially immigrants. Though the relationship between social capital and well-being is well documented, these findings suggest an increased importance for immigrants. Employing mechanisms for increasing social connections has the potential to benefit the health and well-being of this population. Future research should focus on better understanding factors that are more relevant to well-being for native-born Latinas and children of Latinas.