A Biocultural Perspective on Fictive Kinship in the Andes: Social Support and Women's Immune Function in El Alto, Bolivia


Material Information

A Biocultural Perspective on Fictive Kinship in the Andes: Social Support and Women's Immune Function in El Alto, Bolivia
Series Title:
Medical Anthropology Quarterly (MAQ).
Physical Description:
Journal Article
Hicks, Kathryn
Publication Date:


Subjects / Keywords:
Social support
Immune function


This article examines the influence of emotional and instrumental support on women’s immune function, a biomarker of stress, in the city of El Alto, Bolivia. It tests the prediction that instrumental support is protective of immune function for women living in this marginal environment. Qualitative and quantitative ethnographic methods were employed to assess perceived emotional and instrumental support and common sources of support; multiple linear regression analysis was used to model the relationship between social support and antibodies to the Epstein-Barr virus. These analyses provided no evidence that instrumental social support is related to women’s health, but there is some evidence that emotional support from compadres helps protect immune function
Collected for University of Florida's Institutional Repository by the UFIR Self-Submittal tool. Submitted by Yasemin Akdas.
Publication Status:
General Note:
(Citation/Reference) This is the author's post-print. Please cite the final version of the article, available from the link on the cover page.

Record Information

Source Institution:
University of Florida Institutional Repository
Holding Location:
University of Florida
Rights Management:
All rights reserved by the submitter.
System ID:

This item is only available as the following downloads:

Full Text


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


! Kathryn Hicks Department of Anthropology University of Memphis (kahicks2@memphis.edu) A Biocultural Perspective on Fictive Kinship in the Andes: Social Support and Women s Immune Function in El Alto, Bolivia ______________________________________________________________________________ This article examines the influence of emotional and instrumental support on women's immune function, a biomarker of stress, in the city of El Alto, Bolivia. It tests the predict ion that instrumental support is protective of immune function for women living in this marginal en vironment. Q ualitative and quantitative ethnographic methods were employed to assess perceived emotional and i nstrumental support and common sources of support; multiple linear regression analysis was used to model the relationship between social support a nd antibodies to the Epstein Barr virus. These analyses provided no evidence that instrumental social support is related to women's health, but there is some evidence that emotional support from compadres helps protect immune function. [social support, immune function, Bolivia, biocultural] ______________________________________________________________________________ Early investigations of the relationship between aspects of social relationships and health revealed that social isolation is a powerful predictor of mortality (House et al. 1988) The concept of social support emerged from effo rts to clarify the characteristics of social interaction responsible for this pattern partic ularly the positive, functional aspects of social ties (Uchino et al. 1996) Cobb (1976:300) defines social support as the perception that one "is cared for and loved, esteemed, and a member of a network of mutual obligations." Social support is often classified based on the type of aid it confers : Franks and colleagues (2004) describe emotional social support as a relationship that reinforce s feelings that one is valued and instrumental social supp ort as aid in labor or money. A number of scholars have suggested that the most powerful influence on physiology including endocrine cardiovascular (e.g. Eisenberger et al. 2007) and immune function (e.g. Cohen et al. 1997) comes from the perception that support i s available, rather than its actual mobilization (House et al. 1988) Scholars further distinguish between the direct and buffering eff ects where support is seen to alter an individua l's appraisal of, or lessen their physiologi cal reaction to, a stressful circumstance (Cohen and Wills 1985) There are a number of val idated instruments for assessing social support that measure both perceived general support and levels of satisfaction with this support for example (Doeglas et al. 1996) and access to support with regard to specific problems (Winemiller et al. 1993) These are important tools f or the purposes of generalizability but they often make assumptions about social rel ationships (e.g., that all sources are equal and that all support is positive ) (Berkman et al. 2000) Relatively f ew studies have explore d the importance of how cultural context influences norms of social support or have examine d these relationships in non Western settings (Berkman et al. 2000; Dressler and Bindon 1997 2000; Jacobson 1987) Anthropologists have argued that c ulture specifies the norms a ssociated with various classes of social ties and under what general circumstances these relationships can be appropriately drawn upon, which can influence the costs and benefits of doing so ( Dressler and Bindon 2000 ; Dressler et al. 1986; Dressler and Santos 2000) In particular, cultural classifications of kinship and fictive kinship serve, in part, to define expectations associated with


! # specific catego ries of relationship, including expectations of support (Jacobson 1987) These relationships are also not static; the meaning attached to these concepts is influenced by a changing political economic context (Wutich 2011) Ethnographic explorations of the context of social support are important for clarifying the dialectical relationship between changing norms and individual /household strategies and their associatio n with specific health outcomes Several anthropolo gists have relied on ethnographic metho ds to explore cultural no rms of social support and their relationship to health In a qual itative study of the elderly, Jacobson (1987) reported a hierarchy of support expectations, such that individuals expect more support from their female kin, particularly daughters, but excuse those with family obligations of their own. W orking in a community in central Mexico, Dre ssler and colleagues (1986) found that social support from fictive kinship ties, specifically ritual co parents, was associated with lower blood pressure only for men. In addition, while older women and men be nefited from friend support, this was associated with higher blood pressure for younger women. These authors explain this positive association as a result of the fact that women are expected to serve, first and foremost, as a source of support to their own family More recently, Dressler and colleagues (Dressler et al. 1997; Dressler and Bindon 2000) used consensus analysis to examine the cultural dimensions of social support. Work ing in Brazil, and among African Americans, these authors developed cultural models of social support in these communities specifically, a culturally accepted hierarchy of resort with regard to particular problems Cultural consonance, or the degree to whi ch individuals maintained this hierarchy in their perceived support for various problems, was more predictive of blood pressure than overall levels of perceived support, indicating that asking the right person for help is as important as having someone to ask. A number of scholars have investigated the direct effects of Instrument al support In a prospective study among low income Americans, Henly and colleagues (2005) found that individuals with higher levels of perceiv ed support reported lower perceived economic hardship and that an improvement in support over time led to a reduction in the likelihood of experiencing actual material hardship such as hunger or h omeless ness Hadley and colleagues (2007) working with women in rural Tanzania, found that perceived instrumental social support was related to greater food security during the lean season, controlling for household wea lth, and that the relationship was stronger among wealthier households, suggesting they were able to derive gr eater benefits from their relationships. Social support may also be particularly important for households headed by women (Lemke et al. 2003) Kinship in the Andes In many parts of the Andes, inheritance is parallel; husbands inherit land from their father, and women from their mothers, with both partners retaining control of their own resources (Collins 1986) I n other areas, perhaps based on Spanish social norms, inheritance is through the father, and biased toward sons reducing women's status and economic independence (Buechler and Buechler 1971; McEwen 1975) Further, as in other parts of Latin America, women are increasingly likely to head their own households (Staten et al. 1998) Many feminist scholars have questioned the applicability of the household or biological family as a fundamental unit of production and reproduction (Collins 1986; Nash 1993; Van Vleet 2002; Weismantel 1995) For example, among rural Aymara, women and men both retain the rights to their inheritance upon marriage and rely to a large degree on extra household labor for agricultural production (Collins


! $ 1986; Harris 1978) Collins (1986) argues that an increasing emphasis on the nuclear family is evidence of a loss of social resources; w he reas engagement in wage labor increases the difficulty of creating these ties (Collins 1986) Compadrazgo (ri tual co parenthood) has long been an important means of developing extra household ties in the Andes. This is a Spanish term for a custom from Europe, and associated with Catholicism but it reflects practices that were common in the Andes in pre Columbian times (Buechler and Buechler 1971) In this system, children have god parents or padrinos for important events such as baptism, graduation, and marriage, and these padrinos become the co parents, or compadres of the child's parents. Generally, the more equal and important of the relationships is between compadres (Buechler and Buechler 1971) Compadrazgo also plays a role in urban communities (e.g, in structuring obligations relate d to participation in neighborhood festivals ) (Goldstein 2004; Lazar 2008) ; e vangelistas however, a significant and growing minority, generally eschew as pagan communal religious practices like festivals and compadrazgo (Canessa 2000; Van Vleet 2011) A mong other things, the focus of Protestantism on individual salvation over mutual obligation is likely more consistent with participa tion in the market economy and household retrenchment though it may lead to other forms of community building (Canessa 2000; Stoll 1993) El Alto El Alto ( population 800 000) overlooks the colonial city and administrative capital of La Paz, contained within a bowl shaped valley. Early migrants to the city built houses ascending the valley walls, but starting in the 1950s, settlement began to spill out onto the altiplano D eclar ed an independent city in 1988, El Alto was fed initially by min e rs displaced with the closing of government owned mines, and farmworkers freed from haciendas following agricultural reform in 1952 (McEwen 1975; McFarren 1992) It l argely developed during a period of austerity and neoliberal governance starting in the 1980s (Arbona and Kohl 2004; Gill 2000) R esidents have become used to the absence of the stat e and are largely self reliant ; for example, neighborhoods have organiz e d to build their own water and sewer infrastructure (Laurie and Crespo 2007; Revilla 2011) Alte–os demonstrate a high degree of civic participation, with nested organizations from neighborhoods to the Federaci—n de Juntas Vecinales (Federation of Neighborhood Organizations) thro ugh which they engage with the state (Lazar 2008) The city is primarily composed of first or second generation migrants and many retain agricultural plots in and economic relationships with rural communities. Among the f actors that make livelihoods particularly precarious within the city include an absence of formal employment opportunities, a small tax base, state neglect, and an infrastructur e that has not kept pace with the rapid urban growth (Arbona and Kohl 2004; Lazar 2008) This article is based on broader research designed to assess interactions between household composition, access to social support, and women's health in the context of changing social norms and economic hardship. It explore s norms of social support particularly the role of kinship and fictive kinship for women living in El Alto and the relationship between social support and antibodies to the Epstein Barr v irus This analysis tests the predictions that instrumental social support i s a particularly important coping mechanism for women in this marginal urban community that the most important source of support is family and that women benefit from extra household relationships with fictive kin.


! % Methods This study involved a sequential mixed methods approach (Driscoll et al. 2007) using analysis of qualitative data to design structured interviews. Anthropological work on social support indicates that approp riate sources and types of support are highly dependent on cultural and economic context. Anthropologists have used a mix of qualitative and structured ethnographic methods to explore norms of social support, and develop context appropriate measures for qu antitative hypothesis testing (e.g. by asking about problems for which people are likely to turn to others for help and about specific sources of support ) ( Dressler 1991; Dressler et al. 1997; Dressler et al. 1986; Jacobson 1987; Janes and Pawson 1986) Sampling Procedures Participants in this study were drawn from a women's craft cooperative in E l Alto This organization was designed to provide women with regular income generating activities in a setting with few opportunities for formal employment This strategy holds one aspect of socioeconomic status employment constant, in an environment where this is difficult to measure. Women from the cooperative complete most of their work at home but spend one afternoon of the week at the workshop to drop off completed items and pick up supplies and orders. All data collection procedures were carried out by the author in collaboration with an experienced female local research assista nt. The sample includes all members who chose to participate in the study. We completed all the interviews and biological data collection in a separate room in the workshop o n women's normal workdays. T he Northwestern University I nstitutional Review Board and an administrator at Hosp i tal San Gabrial La Paz conducted review s of the research protocol Data Collection During the first phase of the st udy we collected semi structured interviews with six women Specific goals for these interviews included gaining a qualitative understanding of how people conceptualize social support or what kind of language they use to talk about it. During the second phase, w e collected longer semi structured interviews from a group of 29 women (including the same initial six) to assess the utility of possible questions and to explore a broader range of variation in perceptions of and experience with social support. Qualitativ e analysis of these data (Bradley et al. 2007) was used to develop measures of social support, described in more detail below. In the final phase, we conducted structured interviews and biolo gical data co llection among most members of the cooperative ( n =91 a bout 80%) Before this final round of interviews, we spent a week demonstrating biological data collection methods and answering questions about the study. In agreement with leaders of the cooperative, we contributed food to the daily meal shared by women in the workshop and provided individual women with information on their health. This kind of reciprocity was important for building relationships in this setting.


! & Independent V ariables Soc ial Relationships We asked whether participants had compadres, padrinos, a nd family in El Alto or La Paz Instrumental Social Support This is often assessed based on individuals' perception that they have someone to turn to in case of a financial emergency or to borrow money (Hadley et al. 2007; Henly et al. 2005) In this c ase, qualitative analyses of the first two rounds of interviews indicated that the two most common forms of instrumental support discussed by women, both in terms of support they had accessed in the past and support in hypothetical terms, were needing mone y and help with childcare. In the final round of data collection, we asked participants to list the individuals to whom they would turn if they had a financial crunch and needed money and if they needed someone to take care of their children. In each case, the answers were recorded as the number of potential sources of support for each potential problem. Finally, with regard to specific relationships, we asked women whether they thought they c ould ask members of their family, their padrinos, or compadres for money in case of an emergen cy. Emotio nal Social Support As with instrumental support, we asked participants to list people to whom they would turn when they needed emoti onal support or someone who cared for them and would be there to li sten when they had a problem They were also asked if they could turn to members of their family, their padrinos, or compadres for emoti onal support when necessary Control Variables Age and Reproductive Hi story We asked participants to report their age in years and how many children they had born e We also asked whether they were currently pregnant or breastfeeding. Socioeconomic Status (SES) SES has been shown to moderate the relationship between social support and health in some cases (Fagundes et al. 2012) All the participants in this study have the same job, but they were asked whether they had additional employment, about their husband's employment, and their years of education. Goldstein (2004) reports that many Bolivians are reluctant to answer questions about their income and other aspects of SES. On the recommendation of our research assistant and leaders of the cooperative, we did not as k women to report their income or about their access to consumer goods. Instead, we asked whether they had a school, hospital, and health clinic in their neighborhood; whether they had access to water, plumbing, electricity, and a gas stove; how many rooms they had in their house ; and whether they owned, rented, lived with a relative, or paid for their lod gings through anticr’tico (a form of lease where the landlord is lent a lump sum of money over a specified time period) Household C omposition Women may have more or less need for social support depending on their marital status and number of depende nts (Lemke et al. 2003) We collected information on the ages and occupations of all household members and about household headship following State n and colleagues (1998)


! Anthropometrics Weight in kg ( to the nearest 100 g) and body fat (bioelectrical impedence analysis [ BIA ] ) was measured with a Tanita BF 350 analyzer and used to calculate fat free body mass (FFM) in kilograms. C reactive P rotein (CRP) CRP is released by the liver as part of the acute phase response to tissue injury and can increase by a factor of 1 000 in response to a wide range of infections (Black et al. 2004; Mortens en 1994) M arked elevations are the result of acute bacterial or viral infection (Mortensen 1994) CRP concentrations are generally highly skewed and therefore cannot be used in continuous form in multivariate analysis; e levations of >5.0 mg/L have been used as indication of active infection among children and adolescents (McDade 2001 2002) CRP concentration s in mg/L were assayed fr om finger prick blood spots and applied to standardized filter paper (described in more detail below) following procedures outlined by (McDade et al. 2004) Immune Function Blood Spot Collection Finger prick blood spots were collected following McDade and colleagues (2000) One finger was cleaned with an alcohol pad, and a sterile, disposable microlancet was used to make a small puncture. U p to five drops of blood were applied to standardized filter paper commonly used for neonatal screening (#903, Schleicher and Schull, Keene, NH). The samples were covered and allowed to dry overnight, and then transported to the Instituto Boliviano de Biolog’a de Al tura to be stored at 18C. I carried the samples to the Laboratory for Human Biology Research at Northwestern University and stored them at 30C prior to analysis. The samples were exposed t o ambient temperatures for fewer than three days, within the lim its necessary to m aintain sample integrity (McDade et al. 2004) Epstein Barr Virus Antibodies (EBV antibodies ) EBV is a herpes virus that persists in a latent state throughout the lifetime of infected individuals. In developing countries, virtually 100% of individuals are infected by this virus by early childhood (McDade et al. 2000) Althou g h most healthy individuals are able to successfully suppress viral activity after primary infection, impaired immune function due to stress allows the virus to multiply and results in increased antibody concentrations (Glaser 2005; Kiecolt Glaser et al. 2002) A wide body of evidence supports the use of EBV antibodies a s a sensitive measure of or HPA and sympathetic activation (e.g. Cacioppo et al. 2002; Glaser 2005; Kiecolt Glaser et al. 2002; Yang et al. 2010) Anthropologists have recently use d EBV antibodies to study physiological states related to status incongruity (McDade 2001; Sorensen et al. 2009) discrimination (McClure et al. 2010) and social life in a conservative, war torn country (Panter Brick et al. 2008) Potential confounders of EBV antibody concentrations include current infection and nutritional status (Chandra 1997; Kiecolt Glaser and Glaser 1988) We tested percent body fat and C reactive protein as potential confounders Data Analysis Laboratory Analysis I analyzed b lood spot samples at the Laboratory for Human Biology Research at Northwestern University using enzyme linked immunosorbent assay (ELISA) according to previously validated protocols.


! ( Assay of EBV A ntibodies The following protocol was developed and validated by McDade and colleagues (2000) to use a commercially available assay kit (No. 7590, DiaSorin Corporation, Stillwater, MN), to measure p18 VCA IgG antibodies to EBV. One day prior to analysis, a 2.5 mm hole punch wa s used to remove a sample of each blood spot, and these were eluted in 250 L of diluent buffer, overnight at room temperature. One hundred L of eluate wa s added to each well, and the assay completed following kit instructions. The kit includes four levels of p18 VCA IgG standard, and th ree levels of control stabilized human sera. Four additional levels of control in whole blood we re added to filter paper, dried overnight, and stored at 23 C. Standard concentrations, supplied by the manufacturer, we re used to derive standard ELISA scores from absorbance values. Note : the unit of measure for this variable is ELISA units, rather than units of concentration per unit volume of blood. Assay of CRP This protocol was developed and validated (and is explained in detail) by McDade and colleagues (2004) as a high sensitivity assay of CRP concentrations Statistical Analysis All analyses were performed usi ng STATA 8.0 (Stata Corporation, College Station, T X ). All continuous variables were tested for normality by visual assessment of the distribution and statistical tests ( s kewenes and kurtosis, and Shapiro Wilk tests for normality). Where possible non normally distributed variables (age BMI) were log transformed to preserve continuous variation. CRP was dichotomized at three mg/L, with approximately 10% of the sample showing equal or higher levels of elevation. L east squares m ultiple line ar regres sion analyses w ere used to explore the relat ionship between soci al support and EBV antibodies controlling for appropriate conf ounders with a significance level set at alpha < 0.05. With regard to SES, ratios of the number of economically active individua ls and room s in the house were divided by the number of people in the household to create a ratio, and were then log transformed to normalize their distributions. Neighborhood ( number of school s health clinic s and hospital s ) and household characteristics ( presence of running water, plumbing, electricity, gas stove) were added and then dichotomized into two groups of higher and lower SES. Education was dichotomized using a median split, and type of housing was modeled as a dummy comparing renters and women living at home to women who own their own house. A ccess to social support was modeled in two stages. First, dummy variables were created comparing women with zero, one and two or more sources of either emotional support or instrumental support (source o f money) Second, dummy variables were created comparing women with and without either instrumental or emotional support from family, compadres and padrinos. Finally, the relationship with EBV antibodies and a number of potential confounders were tested to determine which variables to include as controls in the final models. Because of the small sample size, variables that had no significant relationship to this outcome were excluded


! ) Results In the preliminary interviews, we asked the participants what kinds of problems they typically need ed help to solve. The most common answers among the sample of 29 women who participated in the first two sets of interviews (some of them twice) were help with an illness, needing money, and needing advice regarding family or interpersonal conflict. We also discussed with participants the problems that cause d them the most worry or preoccupation. The most common problems were needing money, provid ing for their children's education and future (both mentioned by almost everyone), having a stable source of employment, and family and interpersonal conflict, particularly with a spouse or mother in law. Of the sample of 91 women, most ha d family in the area and three quarters ha d compadres and/or padrinos. Table 1 displays the number and percentage of women with compadres and padrinos with perceived instrumental or emotional support from these relationships. In each case emotional wa s more common than i nstrumental support and support from kin more common tha n from fictive kin. The mean number of family members in the area was 10.1. When asked how many sources of instrumental and emotional support they could turn to (general perceived support), participa nts reported an average of 0.9 (SD 0.6, min 0, max 2) and 1.5 (SD 0.9, min 0, max 5) respective ly. Twenty one (23%) women reported no source of in strumental support, while only eight (9%) reported no source of emotional support. Table 1. Number and percentage of women with compadres, padrinos, and family in La Paz El Alto; average number of family in the area ; and number and percentage of women perceiving support from these relationships. N (%) Mean (SD) IS* N (%) ES* N (%) Family 84 (92) 10.1 (9.2) 24 (26) 50 (55) Compadres 66 (73) 11 (17) 14 (21) Padrinos 68 (75) 8 (12) 23 (34) *IS: instrumental support, ES: emotional support In the initial ethnographic interviews, we asked women to discuss their compadrazgo relationships particularly whether they perceived any benefits associated with these connections. Most of the women reported that they either did not have compadres or saw them only rarely (e.g. once a year at Christmas ) Although one woman felt that this relationship was simply a formality associated with baptism, several others conveyed that they shared an affectionate and mutually supportive relationship with their comp adres Those who described a benefit other than friendship and moral support from compadrazgo ties highlighted the importance of the provision of counsel and direction to their children by their baptismal compadres In the larger sample, 66 of the 91 wome n had compadres The probability of having compadres went up with the number of children rather than women's age (Wilcoxon rank sum, z = 3.6, p =0.000). Women self identified as both Catho lic and Protestant participated in these relationships Most of these ritual ties were associated with baptism, but there were others, for ex ample, for a child's graduation. Fourteen women described their relationship with their compadres as emotionally supportive, and 23 described their padrinos as such Eleven women report ed that they receive d some kind of economic or instrumental support from their compadres ; only eight said the same of their padrinos.


! Table 2 presents perceived support, or who women would turn to if they needed immediate help regarding several problems. A fair number of women reported that they would have no one to turn to if they needed money. F or most problems, participants were likely to turn to family first, follo wed by friends. P adrinos we re a potential source of support for marital conflict, but we r e otherwise abs ent. A comparison of Tables 1 and 2 reveals that the different questions yielded slightly different results. When asked if they could ask their compadres or padrinos for instrumental or emotional support, many women said yes but were less likely to list them as a first or second order source of support with regard to specific problems like needing a loan Finally, women were less likely to turn to any of these sources with regard to illness and needing a loan and were more likely to turn to colleagues or supervisors for work related problems than friends, family, neighbors or fictive kin. Table 2 P erceived support, including sources of first and second resort. Problem No one N (%) Family N (%) Friend N (%) Neighbor N (%) Compadres N (%) Padrinos N (%) Other N (%) Loan 1 st 23 (25) 39 (43) 6 (7) 1 (1) 0 0 22 (24) Loan 2 nd ( N =11) 3 (27) 3 (27) 0 3 (27) 0 2 (18 ) Childcare 1 st 8 (11) 43 (57) 5 (7) 4 (5) 0 0 16 (21) Childcare 2 nd ( N =10) 2 (20) 3 (30) 4 (40) 1 (10) 0 0 Conflict w Husband 1 st 11 (15) 28 (39) 11 (15) 0 0 13 (18) 10 (8) Conflict w Husband 2 nd 3 (23) 1 (8) 1 (8) 0 5 (38) 3 (22) At work 1 st 5 (6) 23 (27) 9 (11) 0 1 (1) 0 48 (53) At work 2 nd ( N =14) 1 (7) 4 (29) 0 1 (7) 0 8 (51) Illness 0 21 (23) 2 (2) 1 (1) 1 (1) 0 66 (73) Table 3 shows that EBV antibodies r o se with age and the association between these variables was significant In bivariate analyses, EBV antibodies w ere positively associated with neighborhood SES and marginally positively related to household SES, but this second predictor dropped out of significance when modeled with age It was no t related to years of education or the ratio of household workers to total household members. Nor was it related to a summary variable based on a combination of all of these different measures. Since the goal was to control for SES, only neighborhood SES was included in the final models. Finally, EB V antibody concentration was not related to household size or composition. Table 3 Geometric means and 95% confidence intervals for EBV (ELISA units) by age category. Age N Mean 95% CI 18 29 36 128.6 104.2 158.7 30 39 29 130.2 103.2 164.4 40 49 12 201.0 163.7 246.8 50 75 14 169.8 122.7 235.1 Total 91 142.4 125.8 161.3 Percent body fat wa s higher among older women and lower among women who rent ed rather than own ed their homes. CRP wa s significantly higher only among women with higher


! "+ body fat. There wa s no association between EBV antibody and CRP concentrations over three mg/L. For this reason, this variable was not included in the final models. Although BMI was not a significant predictor of EBV antibody concentrations in the bivariate associations, there was a significant association when age wa s added to the regression model. The first relationships tested were between overall perception of access to emotiona l and instrumental social support and EBV antibodies (results not shown) in two separate models Both of these predictors were negatively related to EBV antibodies, but neither relationship was significant. qq A second set of analyse s tested perceived availability o f emotional or instrumental support from kin, compadres and padrinos. There was no asso cia tion between EBV antibodies and perceiving either form of support fr om kin or padrinos or instrumental support from compadres Table 4 displays multiple linear regression models of the relationship between EBV antibodies and emotional support from compadres W omen with emotionally supportive compadres ha d significantly lower EBV antibody concentrations than women without compadre support This relationship is illustrated in F igu re 1. W omen without fictive kin and with unsupportive fictive kin had similar levels of EBV antibodies. Table 4 Regression coefficients and standard errors for multiple linear regression models of the relationship between emotional social support from compadres and EBV antibodies Model 1A Model 1B N =88 N =88 log Age 95.81 (26.23)** Log BMI 110.38 (50.36)* High SES 46.74 (15.70)* Compadre support 52.03 (20.26)* 45.27 (18.75)* R 2 0.06 0.22 p <0.05, ** p <0.001 Discussion This analysis tested the prediction that perceived instrumental social support is negatively related to EBV antibody concentrations, and therefore positively related to immune function. While associ ated in the expected direction, this relationship was not significant suggesting that perceived support does not moderate the effects of stressful economic circumstances on physiology. Further, reports of having drawn on this kind of support in the past were relatively few (Hicks 2008) indicating that these ties are rarely mobilized. This is in keeping with the argument that neoliberal governance and widespread poverty is associated with a weakening of supportive networks and household retrenchment (Gledhill 2005) and consistent wit h the marginal conditions experienced by Alte–os (Gill 2000) This also consistent with research conducted in other settings in Bolivia For example, in Cochabamba Wutich (2011) found less emphasis on compadrazgo or relationships that carry long term and costly reciprocal burdens, and more emphasis on, for example, patron client relationships that can be severed more easily. Similarly, Andean intellectuals and communiti es have reclaimed the ayllu an ancient Andean social unit based on shared territory and kinship to envision a communitarian alternative to Western capitalism (Weismantel 2006) In the case of


! "" the Landless Peasant Movement that is working to build new agricultural communities around social justice and economic cooperation, discourses on reciprocity based on an imagined ayllu routinely come into conflict with the economic needs of individuals and households (Fabricant 2010) Figure 1. EBV antibody co ncentrations and standard deviations for women with and without supportive compadres A second prediction was that family would be the most important source of support. This conclusion is supported by overwhelming emphasis on perceived support from fam ily in response to each type of hypothetical problem and interview responses that family members are the most appropriate source s for most kinds of problems In several interviews, however respondents indicated that they would never want to be obligated t o members of their family for anything and would prefer to seek assistance from institutional sources. Despite this emphasis on family, this analysis provides no support for the conclusion that supportive family relationships, either in terms of instrument al or emotional support, are associated with immune function. That is, this form of support has neither positive nor negative implications for women's h ealth. This contrast s with research conducted in a number of other settings that shows a strong relation ship between social support from family and health outcomes (Dressler et al. 1997; Uchino et al. 1996 ) In Cochabamba, although more residents reported reciprocal relationships with neighbors, the highest frequency of exchanges occurred between family members despite ambivalence toward these obligations expressed by several participants ( Wutich 2011) Wutich (2011) explained this as evidence of a strong sense of social obligation that discouraged people from refusing requests for help from family even when that was not in their short term e conomic interest. The results of this analysis may suggest that the normative expectation of support asso ciated with family ties lower s their influence on individual physiology. Alternatively, the category of family may be too broad to capture variation in experiences with specific kinds of family relationships.


! "# Finally, the only source of emotional support signif icantly related to EBV antibody concentrations was that from compadres. Most of the women in this sample did have compadr es ; however many reported cordial but relatively distant relations with their ritual kin. Among those with no compadres were all the women without children and some but not all Protestants. Despite the hig h frequency of these connections a minority of women reported having relationships that were supportive, and few listed compadres as a first or second order resort to specific problems. Spatial and temporal diversity in social norms is reflected in ethnographic work in the rural Andes This demonstrat es that compadrazgo is an adaptable system and can serve as an important means of reinforcing reciprocal ties between kin and friends (Allen 1988; Bastien 1985; Buechler and Buechler 1971; Nash 1993; Paulson 2006; Weismantel 1988) and as a means for establishing exploitive economic relationships bas ed on race and class (Albro 2000; Mayer 2002; Nash 1993; Paerregaard 1997; Seligmann 2004; Weismantel 1988) In peri urban Cochabamba these relationships are relatively rare, particularly among younger people, and are associated with a low f requency of reciprocal exchange ( Wutich 2011) This analysis similarly does not support compadrazgo as an important means of extra household economic cooperation in this urban setting consistent with the argument that these relationships are deemphasized in conditions of economic hard ship (Wutich 2011) Rather these findings support the conclusion that compadrazgo can serve an important role in formalizing and strengthening supportive relationships in ways that have significance for individ ual physiology. The sampling strategy involved in this research designed to examine interactions between household headship, social support and women's health likely has implications for interpreting the results. Engagement in the cooperative might be expected to prove an important means of networking and developing supportive relationships for women in this sample. For example, friends, family members and neighbors often suggested new members. Although one respondent did stress the importance of close friends she developed in the group, this theme was not emphasized in the qualitative phase of the study. In addition, livelihoods in this city are generally based on family members pursuing multiple forms of informal and uncertain employment (Gill 2000) The relatively stable employment of these women may mean that they experienced more household economic security than ty pical in this city, which may have implications for their social relationships Th is study's pr imary limitation is the difficulty associated with assessing SES in urban Bolivia. The measures used here are similar to those used in the Bolivian census and by local survey researchers precisely because of this difficulty, and residents may not answer th ese questions honestly for fear of being assessed higher taxes (Goldstein 2002) The measure associated with EBV antibodies was related to conditions in the neighborhood, which may be less problematic than characteristics of households. Nevertheless, these models include unmeasured variation in SES An additional limitation is t hat although this study assessed support from kin and fictive kin, participants were not asked whether they had supportive friends, a class of relationship found important in other studies (Dressler et al. 1986) A t hird pr oblem is that the analysis involved several statistical tests, increasing the probability of finding significant relationships by chance.


! "$ Conclusion T his combination of qualitative, quantitative and biological methods provides unique and valuable in sight into women's experience with specific kinds of social support in El Alto. These results suggest that compadrazgo serves an important role in solidifying supportive relationships for some women in a setting where making ends meet is one of many potential daily stressors. The results further support the conclusion that although many women report perceived instrumental support, particularly from family, economic conditions may reduce the value of this as a coping mechanism This is likely a result of changing social norms, including an increasing emphasis on the nuclear family as the primary economic unit and more indiv idualistic survival strategies. Th is household retrenchment may have important implications for women's status and economic independence. More broadly, this research highlights the value of biocu ltural approaches for understanding the cultural and politica l economic context of social relationships. Embodiment in the form of physiological function (Krieger 2005) powerfully illustrate s the costs and consequences of social ties. References Cited Albro, R 2000 Fictive Feasting: Mixing and Parsing Bolivian Popular Sentiment. Anthropology & Humanism 25:142 157. Allen, C J. 1988 The Hold Life H as: Coca and Cultural Identity in an Andean Community. Washington, DC: Smithsonian Institution Press. Arb ona, J. M., and B. Kohl 2004 City Profile: La Paz El Alto. Cities 21:255 265. Bastien, J W 1985 Mountain of the Condor: Metaphor and Ritual in an Andean Ayllu. Prospect Heights, I L : Waveland. Berkman, L. F., T. Glass, I. Brissette, and T. E. Seeman 20 00 From Social Integration to Health: Durkheim in the New Millennium. Soc ial Sci ence & Med icine 51:843 8 57. Black, S I Kushner, and D Samols 2004 C r eactive Protein. Journal of Biological Chemistry 279:48487 48490. Bradley, E H., L A. Curry, and K J. Devers 2007 Qualitative Data Analysis for Health Services Research: Developing Taxonomy, Themes, and Theory. Health Services Research 42:1758 1772. Buechler, H C., and J M Buechler 1971 The Bolivian Aymara. New York: Holt Rinehart and Winston. Cacioppo, J. T., J. K. Kiecolt Glaser, W. B. Malarkey, B. F. Laskowski, L. A. Rozlog, K. M. Poehlmann, M. H. Burleson, and R. Glaser 2002 Autonomic and Glucocorticoid Associations with the Steady s tate Expression of Latent Epstein Barr Virus. Horm ones and Behav ior 42:32 41. Canessa, A 2000 Contesting Hybridity: Evangelistas and Kataristas in Highland Bolivia. Journal of Latin American Studies 32:115 144.


! "% Chandra, R K 1997 Nutrition and the Immune System: An Introduction. Am erican J ournal of Clin ical Nutr ition 66:460S 463S. Cobb, S. 1976 Presidential Address 1976. Social Support as a Moderator of Life Stress. Psychosom atic Med icine 38(5):300 14. Cohen, S., W. J. Doyle, D. P. Skoner, B. S. Rabin, and J. M. Gwaltney Jr 1997 Social Ties and Susceptib ility to the Common Cold. JAMA 277:1940 194 4. Cohen, S., and T. A. Wills 1985 Stress, Social Support, and the Buffering Hypothesis. Psychol ogical Bull etin 98:310 3 57. Collins, J. L. 1986 The Household and Relations of Production in Southern Peru. Comparative Studies in Society and History 28:651 671. Doeglas, D T Suurmeijer, S Brianon, T Moum, B Krol, A Bjelle, R Sanderman, and W van den Heuvel 1996 An International Study on Measuring Social Support: Interactions and Satisfaction. Socia l Science & Medicine 43:1389 1397. Dressler, W W. 1991 Social Support, Lifestyle Incongruity, and Arterial Blood Pressure in a Southern Black Community. Psychosomatic Medicine 53:608 620. Dressler, W. W., M. C. Balieiro, and J. E. Dos Santos 1997 The Cu ltural Construction of Social Support in Brazil: Associations with Health Outcomes. Cult ure Med icine and Psychiatry Psychiatry 21:303 3 35. Dressler, W. W., and J. R. Bindon 1997 Social Status, Social Context, and Arterial Blood Pressure. Am erican J ournal of Phys ical Anthropol ogy 102:55 66. 2000 The Health Consequences of Cultural Consonance: Cultural Dimensions of Lifestyle, Social Support, and Arterial Blood Pressure in an African American Community. American Anthropologist 102:17. Dressler, W. W., A. Ma ta, A. Chavez, F. E. Viteri, and P. Gallagher 1986 Social Support and Arterial Pressure in a Central Mexican Community. Psychosom atic Med icine 48:338 3 50. Dressler, W. W., and J. E. Santos 2000 Social and Cultural Dimensions of Hypertension in Brazil: A Review. Cad ernos de Sa œ de P œ blica 16:303 3 15. Driscoll, D L., A Appiah Yeboah, P Salib, and D J. Rupert 2007 Merging Qualitative and Quantitative Data in Mixed Methods Research: How to a nd Why Not. Ecological and Environmental Anthropology (University of Georgia):18. Eisenberger, N. I., S. E. Taylor, S. L. Gable, C. J. Hilmert, and M. D. Lieberman 2007 Neural Pathways Link Social Support to Attenuated Neuroendocrine Stress Responses. Neu roimage 35:1601 16 12. Fabricant, N. 2010 Between the Romance of Collectivism and the Reality of Individualism: Ayllu Rhetoric in the Landless Peasant Movement. Latin American Perspectives 34:88 107.


! "& Fagundes, C. P., J. M. Bennett, C. M. Alfano, R. Glaser, S. P. Povoski, A. M. Lipari, D. M. Agnese, L. D. Yee, W. E. Carson 3rd, W. B. Farrar, W. B. Malarkey, M. Chen, and J. K. Kiecolt Glaser 2012 Social Support and Socioeconomic Status Interact to Predict Epstein Barr Virus Latency in Women Awaiting Diagnosi s o r Newly Diagnosed with Breast Cancer. Health Psychology: Official Journal of the Division of Health Psychology, American Psychological Association 31:11 19. Franks, H M., T A. Cronan, and K Oliver 2004 Social Support in Women with Fibromyalgia: Is Q uality More Important than Quantity ? Journal of Community Psychology 32:425 438. Gill, L. 2000 Teetering on the Rim: Global Restructuring, Daily Life, and the Armed Retreat of the Bolivian State. New York: Columbia University Press. Glaser, R. 2005 Stres s a ssociated Immune Dysregulation and I ts Importance for Human Health: A Personal History of Psychoneuroimmunology. Brain Behav ior and Immun ity 19:3 11. Gledhill, J. 2005 Citizenship and the Social Geography of Deep Neoliberalization. Anthropologica 47:81 100. Goldstein, D M. 2002 Desconfianza and Problems of Representation in Urban Ethnography. Anthropological Quarterly 75:485 517. 2004 The Spectacular City: Violence and Performance in Urban Bolivia. Durham: Duke University Press. Hadley, C., M. B. Mul der, and E. Fitzherbert 2007 Seasonal Food Insecurity and Perceived Social Support in Rural Tanzania. Public Health Nutr ition 10:544 5 51. Harris, O 1978 Complementarity and Conflict: An Andean View of Women and Men. In Sex and Age as Principles of Socia l Differentiation. J S LaFontaine, ed. P p. 21 40 London: Academic Press. Henly, J R., S K. Danziger, and S Offer 2005 The Contribution of Social Support to the Material Well being of Low income Families. Journal of Marriage an d Family 67:122 140. Hicks, K. 2008 Social Support and Women s Health in El Alto, Bolivia. D octoral dissertation, Northwestern University, Evanston, I llinois. House, J. S., K. R. Landis, and D. Umberson 1988 Social Relationships and Health. Science 241:540 54 5. Jacobson, D 1987 The Cultural Context of Social Support and Support Networks. Medical Anthropology Quarterly 1:42 67. Janes, C. R., and I. G. Pawson 1986 Migration and Biocultural Adaptation: Samoans in California. Soc ial Sci ence & Med icine 22:821 8 34. Kiecolt Glase r, J. K., and R. Glaser


! "' 1988 Methodological Issues in Behavioral Immunology Research with Humans. Brain Behav ior and Immun ity 2:67 78. Kiecolt Glaser, J. K., L. McGuire, T. F. Robles, and R. Glaser 2002 Psychoneuroimmunology and Psychosomatic Medicine: B ack to the Future. Psychosom atic Med icine 64:15 28. Krieger, N 2005 Embodiment: A Conceptual Glossary for Epidemiology. Journal of Epidemiology and Community Health 59:350 355. Laurie, N., and C. Crespo 2007 Deconstructing the Best Case Scenario: Lesson s from Water Politics in La Paz El Alto, Bolivia. Geoforum 38:841 854. Lazar, S. 2008 El Alto, Rebel City: Self and Citizenship in Andean Bolivia. Durham: Duke University Press. Lemke, S., H. H. Vorster, N. S. van Rensburg, and J. Ziche 2003 Empowered Women, Social Networks and the Contribution of Qualitative Research: Broadening O ur Understanding of Underlying Causes for Food and Nutrition Insecurity. Public Health Nutr ition 6:759 7 64. Mayer, E 2002 The Articulated Peasant: Household E conomies in the Andes. Boulder: Westview Press. McClure, H. H., C. R. Martinez, J. J. Snodgrass, J. M. Eddy, R. A. Jimenez, L. E. Isiordia, and T. W. McDade 2010 Discrimination r elated Stress, Blood Pressure and Epstein Barr Virus Antibodies among Latin A merican Immigrants in Oregon, U S Journal of Biosocial Science 42:433 461. McDade, T. W. 2001 Lifestyle Incongruity, Social Integration, and Immune Function in Samoan Adolescents. Soc ial Sci ence & Med icine 53:1351 13 62. 2002 Status Incongruity in Samoan Youth: A Biocultural Analysis of Culture Change, Stress, and Immune Function. Med ical Anthropol ogy Q uarterly 16:123 1 50. McDade, T. W., J. Burhop, and J. Dohnal 2004 High s ensitivity Enzyme Immunoassay for C r eactive Protein in Dried Blood Spots. Clinical Chemistry 50:652 654. McDade, T. W., J. F. Stallings, A. Angold, E. J. Costello, M. Burleson, J. T. Cacioppo, R. Glaser, and C. M. Worthman 2000 Epstein Barr Virus Antibodies in Whole Blood Spots: A Minimally Invasive Method for Assessing an Aspect of Ce ll m ediated Immunity. Psychosom atic Med icine 62:560 56 7. McEwen, W. 1975 Changing Rural Society: A Study of Communities in Bolivia. Cambridge: Cambridge University Press. McFarren, W. 1992 The Politics of Bolivia s Economic Crisis: Survival Strategies of Displaced Tin m ining Households. In Unequal Burden: Economic Crises, Persistent Poverty, and Women s Work. L Berneria and S Feldman, eds. Pp. 131 158. Boulder: Westview Press. Mortensen, R F. 1994 Macrophages and Acute p hase Proteins. Immunology Revi ews 60:143 158. Nash, J C.


! "( 1993 Household Production and Reproduction in the Latin American Economic Crisis: Rethinking Households in Latin America. Anthropology of Work Review 13 :1 2. Paerregaard, K 1997 Linking Separate Worlds: Urban Migrants and Rur al Lives in Peru. Oxford: Berg. Panter Brick, C., M. Eggerman, A. Mojadidi, and T. W. McDade 2008 Social Stressors, Mental Health, and Physiological Stress in an Urban Elite of Young Afghans in Kabul. American Journal of Human Biology: The Official Journa l of the Human Biology Council 20:627 641. Paulson, S 2006 Body, Nation, and Consubstantiation in Bolivian Ritual Meals. American Ethnologist 33: 650 664 Revilla, C. 2011 Understanding the Mobilizations of Octubre 2003, Dynamic Pressures and Shifting Le adership Practices in El Alto. In Remapping Bolivia: Resources, Territory, and Indigeneity in a Plurinational State. N Fabricant and B Gustafson, eds. Pp. 170 204. Santa Fe: SAR Press. Seligmann, L J. 2004 Peruvian Street Lives: Culture, Power, and Economy among Market Women of Cuzco. Urbana: University of Illinois Press. Sorensen, M. V., J. J. Snodgrass, W. R. Leonard, T. W. McDade, L. A. Tarskaya, K. I. Ivanov, V. G. Krivoshapkin, and V. P. Alekseev 2009 Lifestyle Incongruity, Stress and Immune Function in Indigenous Siberians: The Health Impacts of Rapid Social and Economic Change. American Journal of Physical Anthropology 138:62 69. Staten, L K., D L. Dufour, J C. Reina, and G. B. Spurr 1998 Household Headship and Nutritional Status: Female h eaded Versus Male/Dual h eaded Households. American Journal of Human Biology 10:12. Stoll, D 1993 Introduction: Rethinking Protestantism in Latin America. In Rethinking Protestantism in Latin America. V. Garrard Burnett and D. Stoll, eds. Pp. 1 19. Phil adelphia: Temple University Press. Uchino, B. N., J. T. Cacioppo, and J. K. Kiecolt Glaser 1996 The Relationship between Social Support and Physiological Processes: A Review with Emphasis on Underlying Mechanisms and Implications for Health. Psychol og ica l Bull etin 119:488 531. Van Vleet, K E. 2002 The Intimacies of Power: Rethinking Violence and Affinity in the Bolivian Andes. American Ethnologist 29:567 601. 2011 On Devils and the Dissolution of Sociality: Andean Catholics Voicing Ambivalence in Neo liberal Bolivia. Anthropological Quarterly 84:835 864. Weismantel, M J. 1988 Food, Gender, and Poverty in the Ecuadorian Andes. Philadelphia: University of Pennsylvania Press. 1995 Making Kin: Kinship Theory and Zumbagua Adoptions. American Ethnologist 22:685 709.


! ") 2006 Ayllu: Real and Imagined Communities in the Andes. In The Seductions of Community: Emancipations, Oppressions, Quandries. G W. Creed, ed. Pp. 77 100. Santa Fe: SAR Press Winemiller, D. R., M. E. Mitchell, J. Sutliff, and D. J. Cline 19 93 Measurement Strategies in S ocial Support: A Descriptive Review of the Literature. Journal of Clinical Psychology 49:638 6 48. Wutich, A 2011 Shifting Alliances: Reciprocal Relationships d uring Times o f Economic Hardship i n Urban Bolivia Chungara Revi sta d e Antropolog’a Chilena 43 : 123 133. Yang, E. V., J. I. Webster Marketon, M. Chen, K. W. Lo, S. J. Kim, and R. Glaser 2010 Glucocorticoids Activate Epstein Barr Virus Lytic Replication through the Upregulation of Immediate Early BZLF1 Gene Expression. Brain, Behavior, and Immunity 24:1089 1096.