Social Networks and Health: Volume 8

Since the 1970s, the role of social networks in tracing the underlying epidemiology of illness, distress, disease and disability; and, in understanding how people identify and respond to these crises, has become a mainstay of social and behavioral science research. In this piece, we introduce the papers in this volume in three ways. First, we summarize basic tenants of a social network approach, laying out what research has been able to support to date. Second, to avoid the danger of continuing to do more and more network studies that merely replicate what we already know rather than opening up new areas of understanding, we return to the earliest roots of the social network perspective. Third, we use this review in tandem with the volume's papers to mark the cutting-edge boundaries of current research and to identify the issues and questions that remain.

Social network analysis is a collection of methodological tools that provide a variety of measures and models for describing relationships among social actors embedded in networks of social support. Although methodologists have made great advancements in the area of statistical models for networks, little social support research has moved beyond descriptive measurements of network data. Here we review selected social support literature that has begun to utilize a greater variety of network analyses. We describe and synthesize this literature, then discuss other quantitative tools of use in describing and studying support networks. We also review recent developments in theoretical and applied social support research including research on the role of support networks for persons with HIV/AIDS and older adults. We conclude by discussing the most recently developed statistical models for networks, highlighting their use in a variety of contexts.

As the magnitude of the HIV/AIDS pandemic became apparent, attention was directed once again at the control of infectious diseases. Tuberculosis (TB), the focus here, was once thought ‘conquered’ but again has become a matter of concern. Historically, TB control has concentrated on contacts between infectious cases and susceptible persons in their immediate social (often household) environments. The development of DNA strain typing of the bacterium M. tuberculosis, however, provided important new tools from molecular biology. Subsequent work has indicated that transmission could occur more readily than previously believed, i.e. in community settings. Limitations of conventional TB control programs thus were revealed. Here, tools from molecular biology, epidemiology and network analysis are integrated to suggest a new approach to infectious disease control.

This paper argues that case-centered network analysis requires the integration of both quantitative and qualitative analysis. First, we discuss the importance of linkages or connections between individual clients, providers and agencies in forming networks of care. Next, we explore the linkage between the public mental health and HIV treatment systems. A discussion of the potential influence of cases on networks and network change follows. Finally, we argue the value of incorporating a qualitative analytic approach to understanding inter-organizational linkages.

Network researchers have been concerned with evaluating the accuracy of individuals' descriptions of their personal networks for many years. This paper examines the problem of “forgetfulness” and the extent to which it influences the measurement of ego-centric or personal social networks over time. The data come from the Indianapolis Network Mental Health Study (INMHS), a longitudinal study of the social networks of a cohort of individuals newly diagnosed with mental health problems. We focus on 114 people who completed two or more waves of the INMHS and explore the frequency that names mentioned in prior waves were forgotten. The results indicate that the membership of the respondents' networks changed significantly over the three waves; however, less than 5% of the observed change was due to respondents forgetting ties mentioned in prior waves. The vast majority of changes observed in their networks reflected “true” change in the composition of the respondents' social networks. Overall, the results suggest that people with mental health problems can provide descriptions of the changes in their social networks that are reliable and relatively free of recall bias. The implications of these findings for measuring personal networks over time are discussed.

Persons with multiple sexual partners forget a significant proportion of their sexual partners when asked to recall them (Brewer, Garrett & Kulasingam, 1999). Similarly, drug injectors forget a large proportion of the persons with whom they inject drugs. Forgetting of sexual and drug injection partners in contact interviews for partner notification and social network research hinders efforts to understand and control the spread of infectious diseases such as HIV. We address this problem by describing the cognitive structures people use to organize their partners in memory and developing interviewing techniques that exploit these structures to enhance recall of partners. One hundred fifty-six persons presumed to be at high risk for HIV freely recalled their sexual and injection partners. These subjects tended to cluster, or mention successively, partners who interacted with each other, partners with whom they had the same kind of role relationship, and partners with whom they interacted in the same type of location. The pauses between adjacently recalled pairs of partners typically correlated moderately negatively with pairs' perceived level of interaction. Subjects tended to list partners with whom they had more frequent sexual or injection contact earlier in recall than those with whom they had less frequent contact, and they also tended to recall partners in rough forward or reverse chronological order. The four supplementary interviewing techniques we developed for enhancing recall of partners involve cues that correspond to the organizing factors of partner memory we identified (network ties, role relationships, locations, and chronology).

This chapter examines accuracy of data on characteristics and behaviors of network ties by assessing concordance between 327 indexes' reports on their 411 network members and the network members' self-reports. Participants were recruited for an HIV prevention intervention. Almost all (99%) participants had a history of drug use. Concordance rates varied by the type of information reported and raters' characteristics and behaviors. Overall, substantial concordance (kappa = 0.68) was found for reports of network member's HIV status, with sex partners having greater concordance (kappa = 0.80) compared to HIV seronegative and current drug using indexes (kappas ranged 0.55–0.61). Yet, participants generally, and HIV seronegatives especially, tended to underestimate HIV infection of their network members. Reports of network members' current drug use (kappa = 0.45), current injection drug use (kappa = 0.58), and employment (kappa = 0.52) showed moderate concordance, and age showed near perfect concordance (kappa = 0.90). Results suggest that former and current drug users can provide reasonably accurate data about the HIV status and drug use of their network members. Implications to HIV prevention are discussed.

This paper compares older drug users' exposure to HIV infection or to infecting others with the HIV virus to that of their younger counterparts and addresses the extent to which their personal networks, and the macro-networks within which they use drugs, play a role in risk exposure or prevention. We first consider the changing epidemiology of HIV with respect to older adults. Next we utilize two separate but related sets of data to determine if older drug users are at greater risk than their younger counterparts for drug and sex related HIV risk and also whether or not their knowledge base is sufficiently adequate to enable them to make appropriate decisions about HIV related risk avoidance. We then examine the role of social networks in enhancing or reducing risk of HIV infection in older and younger drug users. Finally we consider the position of both sero-positive and sero-negative older drug users in macro-networks of drug users and whether or not their positions increase or reduce their risk of exposure.

In societies where formal channels of diffusion of health risk and prevention-related information and practices are inadequate, informal social networks offer a powerful alternative. This study compares how social networks deal with two types of public health concerns — the prevention of cholera and HIV/AIDS. It is based on data from in-depth interviews and focus groups conducted in Maputo, Mozambique, in June-August of 1998 — in the aftermath of a major cholera epidemic and in the midst of an unraveling catastrophe of HIV/AIDS. For both diseases interaction through social networks compensates for the insufficient and inadequate information distributed through official channels. However, the differences in the latency period, course, and clinical manifestations of the two diseases affect the nature, mechanisms, and forms of social interaction. Because HIV/AIDS remains largely an abstract threat not supported by practical knowledge and everyday experience, HIV/AIDSrelated social interaction transmits mainly rumors and unconfirmed suspicions. In addition, this interaction is constrained by traditional social barriers and distances, such as the ones based on gender and age, which may retard the rise of HIV/AIDS awareness in the community. In contrast, the symptoms of cholera are easily identified and understood, and the generalized threat of the epidemic spawns intensive and widespread interaction both within and across conventional social boundaries. This interaction mobilizes the community against the epidemic and helps individuals to improve the prevention of infection.

This study examines the positive and negative effects on mental health of three types of network members — friends, relatives and partners. It focuses on how the mental health impact of these relationships changes among young adults who are single, cohabiting, married, or divorced. We test the hypotheses that the negative impact on mental health of problematic relationships varies inversely with the degree to which they are voluntary while the positive impact on mental health of supportive relationships varies directly with the degree to which they are voluntary. Data come from a sample of 1257 young adults who were 25–31 years-old in 1992–1994. Mental health status is controlled through the use of a depression measure obtained from the same persons seven years earlier. In general, the findings support both hypotheses. Problems with partners and relatives have the strongest direct relationships with depression while support from friends has the strongest inverse relationship with depression. For this sample of young adults, relationships with friends have a stronger relationship to mental health than relationships with relatives. However, the results vary across marital statuses: friendships are especially important for the mental health of single and divorced people while problems with partners have a greater impact on depression for married than cohabiting people. The implications of the findings for the functions of network members on mental health are discussed.

In the present chapter, we use data from two studies of family caregivers to elderly relatives to investigate whether stressful circumstances interact with experiential similarity to produce gender similarity in patterns of support. Taken together, the set of findings demonstrate that gender is important in explaining patterns of support among family caregivers to the elderly. First, women reported a substantially larger number of associates from whom they received general emotional social support than did men; however, there were no important gender differences in the number of associates who provided emotional or instrumental support specific to caregiving. Second, both quantitative and qualitative analysis revealed that experiential similarity was much less important in explaining men's than women's patterns of support. We suggest that these findings are consistent with the argument that experiential similarity is most important following status transitions when the transition is salient to the individual.

Despite the long-standing interest of sociologists in the impact of social structure on the psychological well-being of individuals, the structural contexts of the support process remain understudied. To begin to fill this gap, some support researchers have used social statuses to tap location in the social structure. Others have analyzed the interpersonal environments in which individuals are embedded by using quasi-network data that describe categories of alters or, less commonly, network data linked to specific alters. We use network data to test models that examine: (1) direct effects of network structure on perceived adequacy of social support; and (2) their direct and indirect effects (through social support) on psychological distress — net of social status effects. Our results suggest that the social network context is more important in the support process than researchers using quasi-network data have concluded.

While social network support has been found to be critical for persons with schizophrenia and other severe mental disorders, these persons are at risk for having smaller, less functional networks than non-clinical populations. Multi-purpose or multiplex network ties are especially beneficial because they are known to be associated with positive outcomes. This study examined which types of factors were related to network multiplexity over a ten-year period in a sample of 234 persons with severe mental illness, three-fourths of whom had schizophrenia-spectrum disorders. We asked whether clinical characteristics, the nature of the sheltered-care social environments in which subjects were living at baseline, and the number of residential care episodes predicted network multiplexity at follow-up. Using three different conceptualizations of network multiplexity as outcomes, we found that, in general, a prior history of long-term institutionalization, social environments rated higher on practical orientation, and fewer residential care episodes over the study period predicted networks deficient in tie multiplexity. Contrary to expectations, psychopathology was related to only one of the multiplexity outcomes.

We present findings from a National Institute on Drug Abuse funded study of drug use and health care. Our data indicate that stigmatization of illicit drug-users in health care settings and health care providers' diagnostic focus on the contributions of drug use to their illnesses delays treatment, lengthens hospitalizations, and increases health care costs. Our findings show that, not unlike their conventional counterparts, drug users consult their peers for advice, self-treatment, referrals to services, or simply to obtain non-judgmental counsel prior to or in lieu of seeking formal care. Such lay consultation processes influence symptom recognition, attribution, and management as well as the process of formal care seeking and its timing. Discussions with drug-using peers shape perceptions of potential problematic health care interactions. In this regard, interviewees and their lay consultants believe that to be labeled as a drug user, either through self-admission or medical diagnosis, compromises the care they receive. We conclude our chapter by discussing appropriate strategies to improve the quality of care and lower the health care costs of treating drug-using patients. Such interventions include acknowledging and supporting existing lay consultation processes, disseminating better health care information through and within drug users' networks, and encouraging health care providers to conduct more holistic evaluations of drug users' health and illnesses.

This paper examines possible explanations for gender differences in depressive symptomatology. Using a life span framework, two explanations of this difference are explored: women's higher likelihood of experiencing life strains and women's more social network relational orientation. Data are drawn from a regional stratified probability sample of 1,436 people ranging in age from 18 to 93. The findings indicate that the magnitude of gender differences in depressive symptoms does change over the life span, that life strains and social network relational orientation do influence the magnitude of gender differences and such influence varies with age. These findings suggest the importance of a life span perspective, as well as life strains and a social network relational orientation for understanding the frequently documented gender difference in depressive symptomatology.