Background
South Asians (individuals from India, Pakistan, Bangladesh, Sri Lanka, Nepal, Bhutan, and Maldives) are the second fastest growing racial/ethnic group in the United States (U.S.), after Latinos [
1]. South Asians have an elevated risk of cardiovascular disease (CVD) and diabetes mellitus, and a higher ischemic heart disease mortality rate than non-Latino Whites and other Asians [
2‐
6]. Despite a growing body of research to explain and reduce these disparities, individual health behavior and clinical risk factors do not fully explain South Asians elevated cardiometabolic risk [
3,
7]; further, individual level prevention interventions have had limited success in this high risk group ([
8‐
10]). Thus, widening inquiry beyond the individual, to the larger social drivers of health,[
11] may offer key insights into the less-well documented social context of health outcomes in at-risk communities.
Prior research suggests that social relationships exert an especially important influence on behaviors and beliefs of South Asians in the U.S. [
12‐
15]. Almost 90% of U.S. South Asians are first generation immigrants who believe that kinship and family ties are paramount, with an emphasis on collectivism, social control, and maintenance of group identity [
12‐
15]. Studies show that South Asians have low levels of physical activity and dietary patterns that contribute to increased cardiometabolic risk [
3,
16,
17]. These behaviors are socially and culturally informed [
18,
19]; yet there is limited understanding of the structure, composition, and function of social network ties among South Asians, the cultural patterning of networks, and how social relationships influence the health of this community. Understanding South Asians’ social lives, their specific functions, and how they are linked to health can help inform effective interpersonal and community-level health interventions for the rapidly growing South Asian community.
Social networks influence health via many mechanisms including: social influence and control; establishment of health beliefs and normative behaviors; feelings of shared identity and belonging; access to resources; and provision of support [
20,
21]. Social networks transmit information, attitudes, and behaviors that determine health outcomes [
20]; early understandings of network processes suggested mechanisms of network influence through social diffusion such that new ideas and behaviors are spread through contact with other people who have already adopted the behavior [
22].
In addition, voluntary affiliation and participation in a community, religious, or social organization has also been shown to influence network composition, social support, and health [
23]. Voluntary membership in organizations has the potential to create more or less diverse social connections and exposure to additional sources of social influence, norms, and support. Thus, measurement of both personal social networks and organizational affiliation may provide novel insights into the social influence processes relevant to health in South Asians.
The overall goal of this research project was to investigate both personal social networks and organizational affiliation in South Asian community, religious, or social organizations among individuals who participated in the Mediators of Atherosclerosis in South Asians Living in America (MASALA) study, a prospective community-based cohort study on CVD risk and incidence in the U.S. South Asian population [
24]. Although a body of research has shown that social networks are associated with health behaviors and outcomes, there is almost nothing known about the personal social networks and organizational affiliations of South Asian immigrants in the U.S. [
25,
26]. This study provides the unique opportunity to advance research on the cultural patterning of social connections in South Asian immigrants and begins to explore the linkages between social networks and health.
Methods
Participants
The Mediators of Atherosclerosis in South Asians Living in America (MASALA) Study is a community-based cohort of South Asians who were free from CVD at baseline. Details regarding recruitment and baseline measurements have been published previously [
24]. Briefly, using surname-based recruitment methods, a community-based sample of 906 South Asians (age range: 40-84 y, 46% women, 98% foreign-born) was recruited between October 2010 and March 2013 from the nine counties of the San Francisco Bay Area and seven census tracts close to Chicago, IL and surrounding suburbs. To be eligible for the baseline MASALA exam, participants had to self-report South Asian ethnicity, be between the ages of 40-84 years inclusive, and be able to speak and/or read English, Hindi or Urdu. Exclusion criteria included a physician diagnosed heart attack, stroke or transient ischemic attack, heart failure, angina, use of nitroglycerin, a history of cardiovascular procedures, current atrial fibrillation, active treatment for cancer, life expectancy < 5 years due to a serious medical illness, impaired cognitive ability, plans to move out of the study region in the next 5 years, living in a nursing home or on a waiting list, and weight > 300 lbs.
Ethics, consent and permissions
The study protocol and procedures were approved by two institutional review boards and all study participants signed informed consent.
Measurement of personal social networks
From 2014-2017, the MASALA study participants were re-enrolled for a 2
nd study visit where personal network characteristics were measured using a standard egocentric approach that examined the network members (alters) reported by the respondent (ego). The surveys were administered in English, Hindi, or Urdu by trained interviewers. To collect egocentric network data on respondents’ close confidantes, interviewers asked respondents to enumerate relevant alters by using a name generator that has been used by the the General Social Survey [
27] (years 1985 and 2004) and the National Social Life, Health, and Aging Project’s (NSHAP) social networks module [
28] to collect data on participants’ core confidants. Interviewers asked participants to list the people with whom they discuss “important matters.” Respondents could name up to ten people; this name generator was selected to identify network “confidants” who have opportunities to exert social influence and normative pressure [
20,
29]. Studies using this approach have yielded important insights about social contacts who are particularly influential [
30‐
32].
Following the enumeration of alters, the interviewers continued with name interpreter questions, which were used to collect information about the first five network members who were listed. Limiting responses to five individuals is a typical approach to reduce respondent burden and the first alters names typically represent the most important individuals within the social network. Name interpreters helped to characterise the type of relationships (e.g., spouse or significant other, friend), sociodemographic characteristics (e.g. age, country of birth), strength of relationship (emotional closeness, frequency of contact), functions of the network members (e.g., social support), discussion topics (e.g., health) and frequency of communication among the five alters.
Measurement of organizational affiliation
South Asians in the U.S. have developed organizational structures that may exert a strong influence on social connections, social support, and cultural beliefs related to health behaviors. We developed a roster of South Asian organizations (religious, social, cultural, community-based organizations) in Chicago and the San Francisco Bay area using key informant input and an iterative approach. Respondents were asked to look through the pre-defined list and circle the organization(s) that they visited within the prior 12 months and to also circle how frequently they visited the organization in the prior 12 months. Respondents could choose multiple organizations if applicable, and they were also given the option to add an organization if it was not on the roster. There was no limit on the number of organizations a person could affiliate with. However, after reporting all their organizational ties, respondents were only asked in more detail about the 6 places they attended most frequently. The study team limited additional responses to the 6 organizations visited most frequently to reduce participant burden and because the organizations visited most frequently were likely to have the greatest influence.
After data collection, the organizations were coded as community-based organizations, spiritual organizations, and places of worship (i.e. temples, churches, and mosques). We used the Internal Revenue Service definition for coding these and distinguishing spiritual organizations from places of organized worship (
https://www.irs.gov/charities-non-profits/churches-religious-organizations). Spiritual organizations focused on religious and spiritual teaching, but were non-denominational, not considered places of worship, and did not provide organized religious services like a church, temple, or mosque.
Social network measures
The MASALA network data are provided in a dyad-level file in which each row contains information about a specific network member for a given respondent (i.e., multiple rows per respondent). The most basic measure of personal network structure was size: the number of names mentioned in response to the first name generator question. The remaining network variables were calculated using information on the first five individuals listed in response to the name generator question.
Density was defined as the number of ties divided by the number of pairs [
33]. A tie was defined as whether or not there was any reported communication between two alters. A fully dense network (d=1.0) indicates that all network members were connected to each other. Densely connected networks typically have a great deal of influence on an individual's behavior but may not offer access to new information or resources, whereas sparsely connected networks may allow for the introduction of new information, but may provide less tangible support [
34].
Network composition variables, which examine characteristics of alters, included the proportion of specific characteristics: South Asian origin, household member, kin, and gender; in a MASALA respondent’s network. We also calculated the average closeness rating (1(low) to 5 (high)) across alters and the volume of contact with alters as contact-days/year based on the participants’ reports of how often they talked to each alter on a 5-point scale, ranging from every day to a few times a year. We calculated the average number of organizational affiliations for each participant.
Assessing social support
The interview asked respondents about emotional social support (e.g. “How often can you share your worries?”) and instrumental social support (e.g. “How often can you rely on this person for help?”) received from the five alters. Responses were categorized as “most of the time,” “sometimes,” and “rarely/never.” We also asked respondents to report on negative social interactions with each alter, such as, “how often does [alter name] make too many emotional or physical demands on you,” and “how often does [alter name] criticize you?”
Health outcomes
Self-rated health was measured by asking participants to rate their health on a continuous scale of 1–10, with 1 being poor health and 10 being excellent health. A categorical measure of self-rated health (excellent, very good, good, fair, and poor) was avoided given there is a wide range of variability in how those of ethnic minority status and foreign born perceive these categories of health [
35].
For each alter listed, the interviewer asked, “Suppose you had a health problem that you were concerned about, or needed to make an important decision about your own medical treatment. How likely is it that you would talk with [name] about this: Would you say very likely, somewhat likely, or not likely”? This question was the same as what was used in NSHAP. We calculated the proportion of the network that the participant was ‘very likely’ to talk with about health.
Respondent sociodemographics and cultural characteristics
Information on participants’ education, income, age, marital status, birthplace, number of years living in the U.S., and religion were collected as previously described [
24]. Cultural characteristics were captured using multiple items. The traditional beliefs scale was a continuous measure asking participants how much they wished South Asian cultural traditions would be practiced in the U.S. Examples of these cultural traditions centered upon food related activities (fasting, eating traditional South Asian foods like chapattis and daal) and partaking in arranged marriage practices [
36]. The scale had a Cronbach’s alpha coefficient of .81 and ranged from 0 to 28 with lower scores reflecting stronger cultural beliefs and higher scores reflecting weaker cultural beliefs. We also asked participants about cultural self-identity by asking them to report on a scale of 1(not at all)-10 (extremely), “How South Asian do you feel,” and “How American do you feel?”
Statistical analysis
We calculated descriptive statistics for all variables of interest, including participant characteristics, network characteristics, alter relationships, alter characteristics, and organizational ties. We examined bivariate associations between participant and network characteristics using Pearson’s correlations and tested whether these correlations were significantly different from 0. Network variables were modeled as continous variables. For presentation in tables, we categorized some continuous participant characteristics (e.g. age, traditional cultural beliefs, education) because it allowed us to clearly (and parsimoniously) describe how network characteristics may differ as a function of participant characteristics. However, when calculating correlations, participant characteristics were analyzed on their original (continuous) scale in order to better preserve relationships between participant characteristics and network characteristics and also to avoid the loss of statistical power that would be the result of collapsing continuous data into discrete categories.
We described alter social support and negative social interactions by their relationship to the ego. We also described participants’ organizational affiliation by organization type and attendance at health-related events at these organizations.
Lastly, we used adjusted linear regression models to examine if network density, closeness with alters, network composition variables, and number of organization affiliations were associated with self-rated health or the proportion of the network with which the ego was “very likely” to discuss his/her health. Each network characteristic was included as a predictor in separate regression models adjusted for age, sex, education, and network size.
All statistical tests were performed using two-sided tests with α = 0.05 and were conducted using SAS, version 9.4 (SAS Institute; Cary, NC).
Discussion
The MASALA social networks ancillary study is the first comprehesive profile of middle- and older-aged South Asian adults’ social networks and association of network characteristics and functions with health. We found that South Asians living in the U.S. have a relatively large confidant network, which is mainly kin-centered and comprised of individuals who are also South Asian. Network characteristics, including size, composition, and density varied by participants’ age, sex, education, income, and cultural factors, suggesting potentially important subgroup differences in social context, which in turn effects sources of influence and support, as well as types of information and resources available to South Asian immigrants. We also found that networks that were more dense, emotionally closer, and had a higher proportion of kin and South Asians, were positively associated with health-related discussions.
Until now, there have been no data on social networks and health in U.S. South Asians, and less than a handful of studies on South Asians in India and the United Kingdom. We found that South Asians reported larger confidant networks (size=5.6, SD=2.6) compared to prior studies in the U.S. and India; however, previous network studies may not have captured the full extent of the personal network because they placed a smaller limit (maximum of 5) on the number of alters reported during the name generator. The present study used a more open-ended approach and allowed respondents to name up to 10 people during the name generator; our study may be more reflective of true network size. [
26,
28,
37] Similar to Latino immigrants in the U.S.[
37], South Asian immigrants appeared to have dense, kin-centered networks that were ethnically homogenous. Eighty percent of network ties among urban Asian Indians were family members [
26]; although this is slightly higher than what we found, both studies demonstrate that family relationships are central to South Asian social networks. Interestingly, in our study, several social and cultural factors were associated with the proportion kin. Higher education and income and a stronger American identity and weaker South Asian identity were associated with a significantly lower proportion of kin in the network, suggesting that socioeconomic status and cultural change influence social network composition. Others have also shown that lower socioeconomic status is associated with a higher proportion of kin in U.S. populations [
33,
38].
We also examined different types social support and interactions among South Asians and found that family members were most common sources of emotional and instrumental support. Interestingly, South Asians appeared to report less emotional support (being able to talk about worries) from network members, including family, than instrumental (being able to rely on when there is a problem). However, the questions used in our survey may not have distingushed between the “availability” of support and “seeking” support; this may be an important distinction since some cross-cultural studies have indicated that Asian Americans tend to seek less support than other racial/ethnic groups [
39]. Our findings deserve further exploration to determine if there are differences in willingness of South Asians to share emotional concerns and seek emotional support compared to other types of social support. Future analyses will examine if there are links between network characteristics, types of social support, and health in South Asians.
The findings that weaker South Asian ethnic identity and weaker traditional cultural beliefs were associated with networks that were less ethnically homogenous, and less dense may have important implications for the norms, health information, and resources available to South Asians. Others have found that ethnocultural identity impacted peer group choices and was associated with personal network composition in immigrant adolescents [
40,
41]. As a next step, we will examine if these network differences are associated with differences in social norms, influence, and support among South Asians immigrants, which in turn could influence health and behavior.
We also found that South Asian immigrants associate with and attend a large number of local South Asian organizations, including community, social, and religious institututions. Participants perceived attendance at these organizations as beneficial to their health and reported socializing with other members, suggesting that South Asian organizations may provide additional sources of support and social connections. In addition, co-participation in South Asian organizations also provides the opportunity for exisiting norms and behaviors to be reinforced. As a next step, we will examine if attendance and affiliation with these organizations influences health behaviors and outcomes [
29], and if co-participation by MASALA study participants in specific organizations is associated with diet, exercise, and obesity [
42]. Our findings are a starting point for determining if South Asian community structures can be leveraged for health interventions. Religious and spiritual organizations were the most commonly reported affiliations, and the potential of these organizations for health promotion and intervention should be explored.
Others have found associations between social networks and self-rated health [
43,
44], with smaller networks being associated with worse health in the elderly and larger, more family-based networks being associated with better self-reported health. In our study, we only found that greater emotional closeness with alters associated with better self-rated health [
21], but did not find associations with network size or proportion kin. Close alters may provide higher levels of social support or access to other resources that improve perceived health. We also showed that network structure and composition, including density, closeness, and proportion kin, were associated with health discussions among South Asians and their network members. Our data lend additional empiric support to prior studies showing that South Asian immigrants rely on close family members for health information and advice [
45]. How these health discussions influence behaviors or health outcomes is an area that has yet to be explored. It would also be intersting to investigate if specific health issues (mental health, domestic violence, sexually transmitted diseases) are as likely to be discussed in South Asian families as general health problems.
Although an egocentric study provides a feasible way of obtaining network information on a large-scale, it has several limitations. Egocentric data is based purely upon the knowledge, reflection, and recall of the ego, which may be inaccurate – especially when describing the relationship between two alters [
46]. Because we did not observe alter’s view of relationships, we were not able to validate the ego self-report. The analysis is still relevant, however, if we consider the fact that an ego’s perception of relationships may be more important than whether or not the perceived relationship is validated by the alter [
47]. We also did not ask participants about affiliations with non-South Asian organizations or about organizations outside their state of residence, thus limiting our understanding of the full range of potential organizational associations South Asians may have, however, the organizations in the roster represent the major organizations in the lives of South Asians In addition, because of the cross-sectional study design, causality cannot be inferred. Lastly, the MASALA study cohort includes middle- and older-aged South Asians, the majority of whom are Asian Indian immigrants with high socieconomic status. Importantly, our response rate to the social networks module was 78%, and non-responders were more likely to have low socieconomic status and be women.
While the sociodemographics of the MASALA cohort are similar to that of the general U.S. Asian Indian population [
48], these results may not be generalizable to all South Asians. In particular, the social networks of U.S.-born South Asians may be quite different from those who immigrated as adults.