NORTHWESTERN UNIVERSITY
Social Support and Women’s Health in El Alto, Bolivia.
A DISSERTATION
SUBMITTED TO THE GRADUATE SCHOOL IN PARTIAL FULFILLMENT OF THE REQUIREMENTS for the degree
DOCTOR OF PHILOSOPHY
Field of Anthropology By Kathryn Ann Hicks
EVANSTON, ILLINOIS June 2008
2 Acknowledgements
There are many people who supported me during this process, helping me stay sane and challenging me to push myself in ways I would not have otherwise.
I would like to begin by thanking the members of my dissertation committee for always providing constructive criticism along with unending support and encouragement. I would like to thank my advisor and committee chair Bill Leonard for his financial, logistical, and intellectual support, while always encouraging me to pursue my unique interests; Chris Kuzawa for his great graduate seminars and his always detailed and thoughtful feedback; Thom McDade for his big picture advice, and instruction and support with laboratory analysis, and; Mary
Weismantel through her suggestions for readings and contacts, and our many discussions – for helping me to be, first and foremost, an anthropologist.
There are many people whose support not only made my dissertation research possible, but also a great experience. I would like to thank Virginia Vitzthum for giving me my first opportunity to work in Bolivia, and advising me on this project. I would also like to thank Drs.
Hilde Spielvogel and Lieselotte Barragan (Hospital San Gabriel), Marina Claros and Mimi
Woodbridge for their helpful advice and logistical support in Bolivia. Finally I would not have been able to carry out any of my data collection without the dedicated support, friendship and expertise of Esperanza Caceres, who always treated me like one of the family.
My fellow graduate students have played an important role in my intellectual development, particularly my colleagues in the Laboratory for Human Biology Research. I would especially like to thank Elizabeth Sweet, Nicole Fabricant and Colleen Nyberg for their thoughtful feedback on written work, and for discussions which helped me develop my ideas at each stage of the research process.
3 My family has been there for me every step of the way, cheering me on during my successes, listening sympathetically during setbacks. They have helped me with everything from copy editing to moving boxes. My mother Gail, brother Adam, Aunt Alison, and grandmother
Sara Jean, in particular, have been instrumental to my success in completing this dissertation.
Finally, I would like to sincerely thank the women who participated in this study for welcoming me to their place of work, sharing their stories, and against all odds, trying to teach me to knit.
4 Table of contents
Acknowledgements ...... 2 List of Tables...... 6 List of Figures ...... 8 Abstract ...... 9 Chapter 1: Introduction and theoretical background...... 11 Introduction ...... 11 Biocultural approaches in biological and medical anthropology...... 14 Political economy and health ...... 19 Contingency in human evolution ...... 26 Political economic approaches in human biology ...... 34 Neoliberalism in Bolivia ...... 38 Political economy and women’s health in El Alto...... 52 Chapter 2: Urbanization and human biology in the Andes ...... 55 Introduction ...... 55 Human biology research in the Andes ...... 56 Urbanization and health ...... 66 El Alto ...... 76 Conclusion...... 82 Chapter 3: Social relationships and health ...... 84 Introduction ...... 84 Social support...... 84 Stress, adaptation and sociality ...... 87 Social support and neuroendocrine function...... 92 Social support and cardiovascular function ...... 96 Social support and immune function...... 103 Social support and nutritional status ...... 111 Social support, measurement and culture...... 114 Kinship, fictive kinship and gender relations in the Andes...... 120 Conclusion...... 126 Chapter 4: Study design and methods...... 127 Introduction ...... 127 Sampling procedures...... 127 Ethnographic methods...... 129 Biological measures ...... 133 Data analysis ...... 136 Chapter 5: Household Structure, Social Relations and Health Status in El Alto...... 139 Introduction ...... 139 Instrumental variables ...... 139 Control variables ...... 146 Biological measures ...... 147 Hematological variables...... 152 Discussion ...... 155 Conclusion...... 159 Chapter 6: Household composition, social support and health ...... 160
5 Introduction ...... 160 Base regression models ...... 162 Household composition...... 163 Social support...... 164 Discussion ...... 170 Conclusion...... 176 Chapter 7: Conclusions ...... 178 Introduction ...... 178 Limitations ...... 178 Conclusion...... 179 References ...... 181 Appendix A: structured interview...... 227 Appendix B: biological data form...... 231
6 List of Tables
Table Page
4.1. Schedule of fieldwork for February through November 2006………………………129 5.1. Number of total individuals and economically active individuals by women’s age...139 5.2. Household headship……………………………………………………………….....140 5.3. 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 receiving support from these relationships…………………………………………..141 5.4. Number of women who needed and received help for specific problems; number and percentage of women needing help who received it, and; number and percentage turning to specific sources of help…………………………………………………...143 5.5. Current, perceived support, including sources of first and second resort……………143 5.6. Neighborhood and household resources……………………………………………..144 5.7. Geometric means and confidence intervals for education, rooms in the house, and ratio of rooms and workers to total number of household residents………………………145 5.8. Husband’s occupation………………………………………………………………..145 5.9. Frequencies of reported illnesses…………………………………………………….146 5.10. Types of consultation for women reporting illness within the last two weeks………146 5.11. Types of treatment for women reporting illness within the last two weeks…………146 5.12. Church based social support and geometric means for attendance, age and number of rooms in the residence by church……………………………………………………147 5.13. Anthropometric measurements by age category…………………………………….147 5.14. Skinfold measurements by age category…………………………………………….148 5.15. A comparison of cutoffs using BMI and BIA……………………………………….149 5.16. Spearman’s correlation matrix for anthropometric indicators…………………….…151 5.17. Anthropometric measures by housing category……………………………………..151 5.18. Anthropometric indicators by neighborhood and household SES…………………..151 5.19. Anthropometrics by reproductive status……………………………………………..152 5.20. Geometric means and 95% confidence intervals for hematological variables, by age category………………………………………………………………………………152 5.21. Spearman’s correlations for hematological variables………………………………..153 5.22. Hematological variables by residence type………………………………………….154 5.23. Hematological variables by neighborhood and household SES……………………..154 5.24. Hematological variables by reproductive status……………………………………..154 5.25. Hematological variables by respiratory infection status……………………………..155 6.1. Base multiple linear regression models, with regression coefficients and standard errors, for body composition, inflammation and immune function………………….162 6.2. Regression coefficients and standard errors for models including household composition…………………………………………………………………………..164 6.3. Regression coefficients and standard errors for selected multiple linear regression models of the relationship between social support and percent body fat…………....165 6.4. Regression coefficients and standard errors for selected multiple linear regression models of the relationship between social support and EBV………………………..168
7 6.5. Regression coefficients and standard deviations for selected models of the relationship between social support and log CRP………………………………………………...169
8
List of Figures
Figure Page
1.1. Relief map of Bolivia………………………………………………………………….39 2.1. Map of El Alto and La Paz, including major route in and out of the cities…………...77 3.1. Physiological pathways of the hypothalamic pituitary adrenal, and sympathetic adrenomedullary axes, linking perceived social support to endocrine, cardiovascular and immune function………………………………………………………………….92 3.2. Physiological pathways linking social support, shown here as a buffer of stress, and the outcomes of this study: body composition, inflammation, and immune function…...113 5.1. Relationship between percent body fat and BMI…………………………………….148 5.2 A comparison of cutoffs using BMI and BIA………………………………………..150 6.1. Mean percent body fat, and standard error, for women with and without support from padrinos ……………………………………………………………………………….166 6.2. Mean percent body fat, and standard error, for women with and without perceived access to financial support……………………………………………………………167 6.3. Mean percent body fat, and standard deviation, for women with and without perceived access to childcare……………………………………………………………………167 6.4. EBV antibody concentrations, and standard errors, for women with and without supportive compadres ………………………………………………………………...169 6.5. CRP concentrations, and standard errors, among women with and without perceived instrumental support (MHH and DHH)………………………………………………170
9
Abstract
Social support and women’s health in El Alto, Bolivia.
Kathryn Hicks
As in other parts of the global south, economic difficulties in rural areas of Bolivia have forced many of Bolivia’s indigenous people to migrate to urban areas such as El Alto, where formal employment and access to services are limited. Women are particularly vulnerable as they must balance economic activities with childcare, and are increasingly likely to head their own household. Social support is positively related to a number of aspects of health, but work on this topic has tended not to explicitly consider the larger scale cultural, political and economic contexts which may affect this relationship. This dissertation examines the importance of social support for women’s health, in this marginal urban setting, with a focus on cultural definitions of social support, and prevailing political economic conditions in Bolivia. The specific objectives are to explore the cultural context of social support, to determine whether instrumental or economic social support is particularly important in this setting, and to determine whether instrumental support is a stronger predictor of health for women who head their own household.
This project was carried out with women working in a knitting cooperative (N=91), and uses a mix of qualitative and quantitative ethnographic methods to examine relationships between emotional and instrumental social support and body composition, inflammation (C reactive protein) and immune function (antibodies to the Epstein Barr Virus).
The findings of this study include relatively high levels of overweight and obesity, consistent with other studies, suggesting that overnutrition is becoming as much of a problem as undernutrition in Bolivia. Instrumental support is relatively less common than emotional support
10 in this sample, but is positively related to percent body fat, indicating that it may help improve food security. Although few women report strong relationships with fictive kin, emotional support from these relationships is positively related to immune function, suggesting that individuals able to maintain these ties experience material health benefits. Finally, there is no evidence that social support interacts either with socioeconomic status or household composition in predicting health outcomes, suggesting that social support does not serve as a means for coping with economic inequality for women in El Alto.
11 Chapter 1: Introduction and theoretical background
Introduction
After spending several weeks in a rural community, I returned to the site of much of my dissertation research, a woman’s knitting cooperative, to find a higher level of activity than normal. Upon opening the front door my research assistant and I were greeted by the smell of fresh paint, and climbed the stairs to find workers laying linoleum in the room used as a play area for children. The main work area, filled with small groups of women chatting quietly and knitting, was unusually busy with children chasing each other between the chairs. Shortly after we arrived, the director of the cooperative came up the stairs with another gringa in tow, and excitedly pulled us into the office for an impromptu meeting. The college language teacher had raised several thousand dollars in her home country and was placed by a local NGO with the cooperative. In collaboration with the Bolivian director, she had decided to use her money to renovate the children’s playroom, and hire a teacher to deliver lessons while the mothers were working. They also hoped to pay a local doctor to do checkups of all the children and wanted our help determining the best way to go about this.
After our formal discussion, in which my research assistant and I agreed to talk with our contacts in the medical community, the volunteer and I spoke for the first time in English. She had raised several thousand dollars and come to Bolivia because she was deeply concerned about poverty in the global South and felt impelled to do something practical. She was determined to spend her money wisely: she didn’t want to simply hand it over to be wasted, but rather wanted to oversee some specific projects that she thought would be particularly effective. She confided that if she could accomplish only a few small goals, she wanted to explain to women the importance of playing with their children in order to stimulate their development, and the
12 importance of proper hygiene for preventing the spread of infections. When I asked her how she could know anything about the general quality of women’s interaction with their children, having only been in the country for a few days, she laughed dismissively, and went out to mingle with the knitters.
During my time in Bolivia, this was only one of several encounters I had with caring, well intentioned and eager volunteers, NGO workers and foreign government staff. In each case, their approach to development work was informed by Western cultural assumptions about the causes of poverty in countries in the global south, reinforced through conversation with local elites. Recurring themes included the inability of indigenous people to allocate their resources appropriately, their vulnerability to exploitation by radical political parties, and their lack of knowledge regarding issues of health, safety and hygiene. In perhaps the most alarming case, a volunteer English teacher at a rural college described to me her genuine questioning of the
“values” of students who chose to purchase prepared food from a local vendor, presumably in contrast with those who ate at home or the school cafeteria. Putting aside any potential differences in cost or nutritional value, her comment assumes an inherent authority to comment on the small scale financial transactions of poor indigenous people.
Most Westerners who come to Bolivia for any length of time, it seems, are missionaries.
Whether they have come to promote a particular religion, or in the case of many development workers, market fundamentalism and “democracy,” many foreigners assume that the laudable goal of improving health and quality of life for people in developing nations is a question of changing the cultural beliefs and behaviors of individuals to more closely match their own. As argued by David Harvey (2005), the neoliberal model of development is so fully embedded in our cultural narratives about economic inequality that few of us stop to think about the
13 contradictions. Rather than acknowledging the wealth disparities directly created and reinforced by our global economic system, poverty, and all the problems that come with it, are explained as the aggregate of individual failures to properly apply oneself. Nowhere is this more evident than in discussions of how to improve racial and economic health inequalities.
As a discipline, anthropology has two important strengths that run counter to this cultural tendency toward naturalizing poverty. The first is our emphasis on cultural relativity, which suggests that we cannot understand people’s decisions, or even their biology, without first immersing ourselves in, and working to understand, their cultural context. The second is a holism that considers a broad range of local and large scale processes as potential, and likely interacting, influences of the human condition. Biological anthropology thus holds a unique power to provide a contextualized and nuanced view of human variation, rarely available from the perspectives of public health or biomedical sciences. To do this, however; we must incorporate appropriate theory and methods from other subdisciplines, particularly cultural anthropology.
My dissertation research was designed to explore the importance of social support for multiple aspects of women’s health in El Alto, Bolivia, with a particular focus on interactions between social support and household composition in predicting health. Though much of the work demonstrating the effects of social relationships for health has come from other disciplines, anthropologists have been at the forefront of research demonstrating the importance of cultural and socioeconomic context for structuring the costs and benefits of social relationships (Dressler and others 1997; Dressler and others 1986; Jacobson 1987; Janes and Pawson 1986). In the following discussion I will argue that, as part of a larger biocultural framework, an understanding of both contemporary and historical political economy is critical to explaining human biological
14 variation both within and across populations. Though much work remains to be done in integrating political economy with ecological approaches, this dissertation represents, for me, a first step in moving toward a political economy of human biology.
In the following discussion, I will examine biocultural approaches to understanding human variation, and particularly, review arguments for and against the incorporation of a more explicitly political economic focus. I will finish with a consideration of the specific political economic context of Bolivia, and how this influences the objectives of this study.
Biocultural approaches in biological and medical anthropology
Wolf laments the current state of affairs in the social sciences in which economic, political and social forces are compartmentalized into different areas of study which “having abandoned a holistic perspective, thus come to resemble the Danae sisters of classical Greek legend, ever condemned to pour water into their separate bottomless containers” (Wolf 1997:
11). The same problem exists, on a smaller scale, within anthropology. Specialization within the discipline has meant little communication and collaboration between cultural and biological anthropologists and in general it has meant a scaling down of the holism upon which anthropology was built (Goodman and Leatherman 1998). While there are benefits to a certain degree of specialization, it is essential to be able link local and small scale processes with regional and global phenomena (Goodman and Leatherman 1998). In addition, it has become increasingly clear that the biological, social and cultural forces involved in the production of health cannot be compartmentalized into different areas of study without the loss of important information on their interactions (Armelagos and others 1992)
On the other hand, Morgan (1998: 413) argues that a large number of leftist scholars have become leery of the study of biological variation because of the “populaces’ willingness to
15 interpret social issues such as race, gender, homosexuality, and poverty through reductionist biological lenses in the service of conservative social agendas.” While this attitude is common among cultural anthropologists, it is precisely because of the tendency for biological data to be misused and misinterpreted that the contribution of anthropology to understanding human variation is so important. Health has profound consequences for human functioning and quality of life, and may be one of the most important means by which social and economic inequalities are embodied. In seeking a nuanced understanding of the relationship between biology and poverty we can demonstrate that reductionist explanations are a function of scholarly approaches, not biology itself. An integrated biological anthropology has the potential to make unique contributions to improving our understanding of health, and in so doing, can also make important contributions to anthropological theory more generally (Dressler, 2001).
A number of investigators have advocated for the adoption of a more sophisticated biocultural approach, which incorporates methods and theory developed in cultural anthropology to better understand human biology (Armelagos and others 1992; Crooks 1999; Dressler 2005;
Dressler and Bindon 2000; Dufour 2006; Goodman and Leatherman 1998; Schell and others
2007). In practice, the term biocultural can obscure as much as it clarifies. On some level, any researcher who studies human variation would be hard pressed to completely ignore the complex social and cultural environments occupied by humans, and investigators working with a diverse range of theory and methods use this term. To further complicate matters, scholars of cultural ecology use this as a means to describe their focus on evolutionary influences of human behavior. None of this would be a problem, per se, if there were not so much conflict within anthropology about what constitutes truly “biocultural” work. The tendency has been to focus too heavily on defending one’s own tack, rather than recognizing the diverse range of approaches
16 available to anthropologists to understand the complexity of human environments and biology.
These different approaches represent fundamentally different research interests, and are in no way mutually exclusive.
The more traditional biocultural approach, as practiced by both medical and biological anthropologists, recognizes individuals as both cultural and biological beings subject to evolutionary processes, and health as a material and measurable phenomenon (Leslie and Little
2003; McElroy 1990; Wiley 1992). To differing degrees, researchers employing this perspective integrate, or consider as variables, cultural, social and economic phenomena thought to be directly associated with health. Investigators informed by this approach tend not to produce research which is explicitly political in its orientation, though they do address issues of material inequality (McElroy 1996; Wiley 1992). Supporters of this body of research rightly suggest that critics often fail to appreciate its strengths. In particular, Leslie and Little (2003) argue that human biology has implications for understanding issues of importance to cultural anthropologists, such as the embodiment of social and economic inequality. The importance of this work in laying a theoretical and methodological foundation for more integrative research cannot be overestimated.
A number of authors have proposed that biological anthropologists need to be more aware of the socially and culturally constructed nature both of biology and perceived illness, following the example of medical anthropology (Armelagos and others 1992; Dressler 2005).
Important considerations include the role of culture in defining illness and appropriate sources of health care, in structuring emotional and physiological reactions to daily events, and in patterning disease risk (Armelagos and others 1992; Dressler and Bindon 2000). Further, cultural anthropologists recognize that human behavior has a performative component, where
17 illness or treatment choices can represent a form of agency through which individuals make claims about their situation or identity (Crandon Malamud 1991). Few biological anthropologists consider this either as a potential influence of self reported health, or as a potential explanation for health related decision making. There is much evidence that culture is a crucial component in explaining variation in human biology (Dressler and Bindon 2000). To look for research tools to address this issue, we need to be conscious of developments in culture theory within cultural anthropology (Dressler 2005).
Critical Medical Anthropologists (CMA) have presented a slightly different critique that is also representative of currents within cultural anthropology more generally. These authors are explicitly critical of the biocultural approach, the use of scientific methods in anthropology, and particularly of the dominant biomedical paradigm in health research. In particular, CMAs have sought to understand how social and cultural biases influence the practice of “objective science” and argue for an increased reflexivity in thinking about how science affects research subjects
(Kielmann 2002; Singer 1998). For example, biomedicine is seen as a form of hegemony imposed by dominant classes, or a means of naturalizing the existing social order (Singer 1998).
Singer (1996), while he supports scientific endeavor and considers human biology important, has argued that the concept of adaptation is no longer useful to biological anthropologists as it obscures the socially constructed nature of the environment, as well as the potential for social inequality to lead directly to differences in survival and reproduction. He argues that human biology cannot be understood without reference to political economy.
Discussion of an improved integration of biology and larger scale forces has also recently become more common among biological anthropologists, following some of the same logic as
CMA. Several authors argue that the dominant adaptationist perspective in biological
18 anthropology, by failing to consider structural inequalities and focusing on individual responses as either ‘adaptive’ or not, tends to blame the victim (Goodman and Leatherman 1998;
Leatherman and Goodman 1997; Levins and Lewontin 1998). Goodman and Leatherman (1998:
19 20) have advocated for the consideration of several important issues for understanding human variation, generally underdeveloped in biological anthropology including: 1) social relationships, particularly with reference to power; 2) links between local and global processes; 3) history and historical contingency; 4) human agency; 5) and the ideologies of scientists and the people they study. Above all, they argue for the use of a political economic perspective, which recognizes the importance of human rights and is more ethically accountable, for understanding human variation and health (Goodman and Leatherman 1998; Leatherman and Goodman 1997).
While there is a general agreement that biocultural models are necessary, there is far less agreement about what these models should include, and how to develop appropriate methods for achieving this end. Cultural anthropologists generally deal with the complexity of culture by using a mix of intensive, qualitative ethnographic methods, the most important of these being participant observation, used to gain, as much as possible, the cultural understanding of an insider. Those cultural anthropologists interested in political economy use this as a means to observe how large scale processes play out in the everyday lives of marginalized groups. A central difficulty for anthropologists interested in explaining variation in health is how to link these more fine grained qualitative methods with statistical models, in a sufficiently representative sample (Crooks 1997). Further, a number of scholars continue to demonstrate reluctance to embrace political economy as a legitimate object of study for biological anthropologists, whether for theoretical or methodological reasons.
19 In the following section I will argue that we need to create a space within biological anthropology for biocultural approaches that explicitly consider the importance of historical political economy. This is in no way to suggest that others need to adopt this approach, but that the sub discipline as a whole will benefit if this is accepted as a valid theoretical contribution.
Political economy and health
Following Dressler’s suggestion that biocultural research needs to keep up with developments in culture theory (Dressler 2005), I suggest that we should begin any discussion of the importance of political economy by looking at the work of cultural anthropologists who take this as their theoretical focus. According to di Leonardo (1998: 77), “’Culture’ is never separate from, and cannot be understood apart from, politics and economy,” whether or not theoretical and methodological issues make this a difficult proposition. For those investigators who accept this argument, any biocultural synthesis that does not take into account politics and economy will necessarily be incomplete.
One of the primary arguments to come from this body of work is that the objects of study for social scientists, “nations,” “societies,” and “cultures” are heuristic devices that actually have as much power to obscure as to illuminate (Wolf 1997). While these terms are useful to investigators for the purposes of discussion, by abstracting them from their global context, the heuristics become reified into theoretical propositions (Wolf 1997). Scholars of political economy reject the notion that cultures or societies constitute bounded entities that can be understood without reference to the larger system in which they are enmeshed (di Leonardo
1998; Roseberry 1982; Wolf 1997). Human groups share complex connections in a process of global, but profoundly uneven development, which is neither an unfolding directed by universal laws of nature, nor a random process (Roseberry 1982; Wolf 1997). Importantly, the ties that
20 link individuals, nations, and regions together are infused with power differentials that structure both local and global interactions. Following this logic, theorists of political economy are profoundly interested in historical processes at the local, regional, and global level which lead to inequality.
Roseberry (1982: 19) in elaborating a theory of culture takes as his starting point the work of Clifford Geertz (1973), who advocated an approach centered on meaning or “the socially constructed understandings of the world in terms of which people act.” Where he differs from Geertz is in seeing culture not as a static product or text, but as both socially constituted and constitutive (Roseberry 1982). That is, culture informs, but is also shaped and changed by action. Further, he sees the creation of meaning as a material process: culture is constructed within relations of inequality where wealth and power determine who has greater access to the means of cultural production, and therefore the creation of meaning (Roseberry 1982). In this way, culture is highly contested by groups with conflicting interests, and can serve simultaneously both to reinforce hegemony, and as a site of resistance. An understanding of culture simply cannot be gained by divorcing cultural production from the dynamics of this contestation, or its material effects.
Contrary to the suggestion of critics, political economic scholars are not exclusively concerned with making generalizations about how global forces play out at the local level. This should be evident in Roseberry’s (1982) conception of culture, discussed above; macro level forces are complex and contradictory, and they are contested in local contexts that are complex and contradictory. Both Wolf (1997) and Roseberry (1982) explicitly reject the idea that circumstances in the periphery can be explained solely as an outcome of forces emanating from the center, and suggest that the role of anthropologists is to understand the importance of these
21 forces and how they manifest at a local level. This theoretical orientation has important implications for the construction of biocultural models. I will explore these implications using the example of evolving studies of culture change and health.
A large number of biological anthropologists are interested in understanding the importance of changing circumstances for local populations, referred to variously as
“modernization,” “westernization,” “acculturation,” or “culture change” in the context of the
Global South (Baker and others 1986; Friedlaender 1987; Shephard and Rode 1996). Broadly, these terms are used to refer to processes of globalization that introduce changes in patterns of labor, diet, lifestyle and social relations that have been well demonstrated to have important influences on human biology (Dressler 1999; Dressler and others 1995; McDade 2002). Early studies based on modernization theory used ecological comparisons, for example, of groups living in rural and urban areas, to draw conclusions about the effects of culture change on health.
Though these studies established the importance of changing circumstances on human biology, the inability to link culture change to health at the individual level precluded fine grained analysis of causal relationships (McDade 2002). With an increasing focus on examining specific aspects of this process, it became increasingly clear that it is complex, with positive and negative effects on human biology. For example, modern medicine and sanitation lead to improvements, while increased sedentism and lower quality diets bring novel health problems (Wirsing 1985).
Dressler and colleagues introduced an important refinement to this approach, in the form of studies of status incongruity and stress, as a means to more fully explore the association between culture change and increasing blood pressure (Dressler 1999; Dressler and Bindon
1997). Following Weber, this theory suggests that with economic development comes an increasing focus on the consumption of Western material goods as markers of social status,
22 without providing the social mobility necessary to put these status items within reach (Dressler and Bindon 1997). Status incongruity comes when individuals struggle to maintain patterns of consumption inconsistent with their economic means, leading to higher levels of psychosocial stress along with the associated downstream health effects (Dressler and Bindon 1997). A number of investigators, working primarily in Samoa, have found support for a relationship between status incongruity and health, and most importantly, have found that this relationship is highly dependent on the local cultural context, and plays out differently for individuals with different social roles (Bindon and others 1997; Bitton and others 2006; Chin Hong and
McGarvey 1996; Dressler and Bindon 1997; Dressler and others 1995; McDade 2001; McDade
2002).
Biological anthropologists have been at the forefront of clarifying the influence of proximate mechanisms on health but have been relatively less critical in their acceptance of the larger social processes causing these lifestyle changes (Goodman and Leatherman 1998; Singer
1996). Studies of status incongruity add an important level of complexity and sophistication to the study of culture change and health, but are not concerned with addressing political economy.
All of these investigators are undoubtedly exploring important processes that have profound implications for human biology but I will argue that “culture change” is not separable from political economy, and that we are missing out on important information by not considering this.
There are two major theoretical problems that remain to be confronted. The first is that models based on modernization theory are inherently teleological, suggesting an evolutionary chain from traditional to modern through which each society will pass, and naturalizing capitalism as the inevitable endpoint (Edelman and Haugerud 2005; Roseberry 1982; Wolf 1997). The second is that terms such as “culture change” imply a neutral process rather than one that is structured by
23 inequality. Both of these underlying assumptions influence the questions that we ask. I will deal with each of these issues in turn.
While a number of authors are impressed by the current scale of global movements of people, ideas and commodities (e.g. Appadurai 1990), others see these movements as somewhat less novel (Edelman and Haugerud 2005; Mintz 1998; Wolf 1997). Mintz (1998: 123), writing about the movement of labor from Europe and Asia to Caribbean sugar plantations in the 19 th century remarks, “Does it not seem, to those who believe globalization is a new phenomenon, that moving a million people a year transoceanically for an entire century is pretty big and pretty global?” The populations studied by anthropologists have long been immersed in global forces, and even relatively isolated populations have felt the effects of it. Characterizing societies as traditional places them outside the flow of history (Wolf 1997). For example, indigenous people in highland Bolivia were conquered first by the Incan empire, and for five hundred years after conquest by Europeans were subject to forced resettlement and labor (Larson 1998), conditions that cannot reasonably be considered traditional. Culture change is not something restricted to the modern age or to certain locations, but a constant, ongoing, and universal process.
Understanding the dynamics of culture change in any one area is fundamentally about understanding historical processes.
Culture change is also not a neutral process, but a highly uneven and contested one. The expansion of European empire was largely fueled by the extraction of resources and labor from the Global South during colonialism, and this continues today under the guise of neoliberal development. As many social movements in Latin America and other regions are increasingly aware, the development promised by the World Bank and the International Monetary Fund is a cover used to deregulate their economies for the benefit of foreign governments and corporations
24 (Harvey 2005; Harvey 2006). For much of the world’s population neoliberal development has meant decreasing wages, minimal spending on public infrastructure and social safety nets, and an increasingly marginal existence (Harvey 2005; Harvey 2006; Klein 2007). Uneven proletarianization, or the segmentation of labor into skilled and unskilled sectors on the basis of race, gender and class, heightens existing inequalities and creates new ones (Roseberry, 1994).
Following Roseberry’s (1982) definition of culture, we can hypothesize that some aspects of cultural construction will be based on the contestation between those who do, and do not, benefit from these processes, and that this contestation will be highly dependent on the local cultural context.
I will use the example of status incongruity to explore just some of the ways in which this definition of culture, that is, a definition that incorporates historical political economy, might add another layer of complexity to a biocultural analysis. The concept of status incongruity makes an important contribution toward understanding the disconnect between the status aspirations of individuals and their ability to achieve these aspirations, and demonstrates the operation of macro level processes on individual health (Dressler and others 1995). A more explicit focus on which factors influence who is more or less likely to be able to achieve their desired level of consumption would likely reveal additional insights. A second issue is that identity formation is complex, and based on more than material consumption, education and occupation. For example, a number of the studies of status incongruity use mass media as a measure of Western patterns of consumption, but media also play an important role in establishing narratives about politics and the economy that are likely to have an important influence on how people internalize their changing circumstances. The degree to which individuals accept, or use cultural expression
25 to challenge dominant cultural narratives may have implications for their consumption patterns, their identity and their health.
More generally, the language we use to talk about health inequalities matters. Using terms such as “modernization” or “acculturation,” even if we are personally critical of the generalizations they suggest, serves to naturalize the neoliberal trend toward increasing inequality. Though a relatively small body of anthropological literature makes it into the public sphere, we have a responsibility to insure that our work reflects the depth and complexity of anthropological knowledge to those outside the field.
My point here is not to argue that biological anthropologists are unaware of these issues or have not sought to deal with them in meaningful ways, but rather that there should be room in the discipline to build upon this existing work. A number of individuals have expressed to me their concern that the only possible contribution of a political economy of human biology is to reinforce the idea that poor people have poor health. Returning to the examples I presented in the introduction, none of the development workers with whom I spoke in Bolivia had any doubt that poverty influences health, and most were eager to try to address the problem. Their lack of understanding of the causes and constraints of poverty, or the complex reciprocal relationship between biology and the environment, meant that they were addressing their efforts in ways that were likely to be less than entirely helpful, and to reinforce the marginalized position of indigenous people. Suggesting that people should wash their hands more often is both insufficient and patronizing if corporate ownership of the water supply means that they cannot afford clean drinking water. If biological anthropologists want to produce research that has practical implications for improving health, we will have to begin to grapple with ways to better address inequality.
26 An important issue for many biological anthropologists is reconciling biocultural with evolutionary approaches in a way that avoids reductionist explanations of the human condition, but recognizes humans as part of the natural world (Leonard 2005; Smith and Thomas 1998). In contrast, many cultural anthropologists are leery of explanations framed in evolutionary terms.
In the following section I will review developments within evolutionary biology that suggest the compatibility of these different approaches.
Contingency in human evolution
Some of the strongest and most productive critiques of reductionist or deterministic models of biological variation have come from within evolutionary biology. A number of authors have struggled to overcome the false dichotomy between genes and environment as determinants of physical form, and to demonstrate the complexity and contingency of developmental and evolutionary processes. Of particular concern to a number of these investigators is counteracting the idea that genes can be described as a blueprint for constructing an organism, and that understanding this blueprint can only come from excluding the “noise” produced by the environment. Much of the impetus for this reassessment comes from an increasing concern for linking the Modern Synthesis with developmental biology, fields which have had relatively little interaction until recent times (Buss 1987; Griffiths and Gray 2001).
Stephen Jay Gould, throughout his career, argued against a teleological view of nature.
He used a reinterpretation of the diverse organisms preserved in the Burgess Shale, a record of the immense diversity of organisms during the Cambrian explosion, to explore the importance of historical contingency in evolution (Gould 1989). While all of the body plans that currently exist in nature were present in this early record, so were a large number no longer represented among extant organisms. He argued that there is no evidence to suggest that any of these forms were
27 any more or less adaptive than others, but that chance may have played an important role in determining which organisms survived a subsequent mass extinction (Gould 1989). If this is the case, then all of the diversity we observe in nature is dependent upon which earlier forms were preserved, and running the process again would result in fundamentally different, though still highly ordered results (Gould 1989). He also suggested that while humans have a tendency to observe patterns and directionality in nature, there is no evidence that evolution demonstrates any form of progress towards increasing complexity (Gould 1996). Throughout the history of life on earth, single celled organisms have dominated, and the development of complexity likely represents the filling of an additional niche that was open (Gould 1996). He argues that Darwin shared this view, and was reluctant to use the term evolution to describe his theory, because it implied an orderly unfolding rather than a fundamentally contingent process (Gould 1996).
In Developmental Systems Theory Oyama and colleagues redefine heredity in terms of the transmission of all developmental conditions, which in addition to genes, includes epigenetic and environmental transmission, as well as their interaction (Jablonka 2001; Oyama 1985;
Oyama and others 2001). In so doing, they reject the idea that genes are the most important cause of biological form, subject to influence by the environment, and the idea that traits can be proportioned into degrees of genetic versus environmental input (Oyama and others 2001).
Rather, “the life cycle of an organism is developmentally constructed, not programmed or preformed,” and ontogeny is a fundamentally interactive process where each source of input is equally important (Oyama and others 2001: 4). Particularly relevant for those who study human variation is the idea that constancy through time cannot be attributed solely to genes, but is itself constructed during development, and that all biological processes are contingent upon complex interactions within constantly changing systems (Oyama 1985).
28 In a similar vein, West Eberhard (2003) suggests that developmental plasticity is a critical component in evolution, as genes are subject to selection only when they are expressed during the process of phenotypic development. In contrast to most introductory evolution texts, she defines evolution, not just in terms of changing gene frequencies over time, but as changing phenotypes. She suggests that development involves a complex interaction between genetic and environmental inputs, with the process of somatic selection serving as a sort of exploratory process through which organisms can adapt to novel circumstances (West Eberhard 2003). An example of this process, which will be explored in a later chapter, is the production of enormous varieties of B cells which are then subject to selection based on the antigens encountered by the organism (McDade and Worthman 1999; West Eberhard 2003). In this way, environmental variation can be an important cause of phenotypes, which if influenced by a genetic component can result in directional evolutionary change. Finally, she suggests that phenotypic plasticity itself would have been selected for at each level of biological organization to allow organisms to adapt to changing conditions (West Eberhard 2003).
Others in the field have taken a Marxist approach to understanding evolutionary biology.
Levins and Lewontin (1985; Lewontin and Levins 2007) support a dialectical view of the natural world, suggesting that biological wholes are made up of interpenetrating parts (themselves made up of dialectical relationships), and that causality runs in multiple directions, and across many levels, at once. They also suggest that change is the inevitable result of these contradictory, interpenetrating forces, and rather than focusing on biological form (a snapshot in time), the process is the appropriate unit of study (Levins and Lewontin 1985; Lewontin and Levins 2007).
These authors reject reductionism, which assumes that understanding parts is sufficient for understanding a whole, as higher levels of organization demonstrate emergent properties based
29 on complex interactions. In a discussion similar to that of Wolf (1997), they suggest that academic specialization which attempts to split the natural world into manageable parts reinforces the tendency toward reductionism, and that the most fruitful research will likely take place at the borders between disciplines (Levins and Lewontin 1985; Levins and Lewontin 1998;
Lewontin and Levins 2007).
There are two important practical implications of this work, which share a surprising degree of overlap with those from political economy, discussed above. The first is that we must avoid teleological thinking, and look to history, or to development, to understand why things are the way they are, and not different. Lewontin and Levins (2007) suggest that this error in logic has produced many ongoing misunderstandings of human biology, for example, the apparent paradox of worsening health accompanying economic development. The second is that contrary to the opinion of many outside of the field, there is no inherent association between the study of biology and reductionism. In practical terms, for example, there is no one to one relationship between genes and traits (West Eberhard 2003). Correlations between genes and biology cannot be taken as evidence that they control these traits, as their expression is fundamentally context dependent (Oyama 1985). This does not mean that anthropologists should avoid studying genes, but rather that we should be far more cautious in our interpretations, and mindful of the complexity and contingency of biological processes.
At the same time, we should continue to appreciate the core strengths of biological anthropology, including a focus on human adaptability. In the same way that evolution, a term synonymous with teleological and progressive change in everyday usage, is often conceived of quite differently by biologists, the concept of adaptation is more complex than sometimes acknowledged. As discussed above, a number of authors have suggested that the term adaptation
30 implies that a trait or behavior is both positive and sufficient to for mitigating environmental stress (Goodman and Leatherman 1998; Leatherman and Goodman 1997; Singer 1996). The danger in this term then would be naturalizing embodied inequalities, and the forces that create them. As I have argued throughout this chapter, the solution to this problem is to focus on social and economic inequality as important aspects of our environment, not to ignore potential evolutionary explanations for human biology. Most evolutionary biologists conceive of adaptation as compromise solutions to competing environmental constraints rather than some optimal solutions without costs or tradeoffs (Mayr 1983).
Adaptation involves the maintenance of homeostasis, or the preservation of a constant internal environment (Frisancho 1993; Slobodkin 1968). For example, diverse traits such as body temperature and calcium concentration in the blood are maintained within precise limits despite fluctuations in the external environment, and failure to preserve these systems leads to death. Other systems may be somewhat more flexible. Allostasis, or the process of maintaining stability through change, has been proposed as a better model for explaining variation in aspects of the autonomic nervous system, the hypothalamic pituitary adrenal axis, the cardiovascular system and the immune system (McEwen 2005a). These systems may react by changing the set point in order to protect the body from stressors, rather than maintaining constant values in the face of changing internal and external conditions (McEwen 2005a). Some authors see allostasis as complimentary to homeostasis (McEwen 2005a), while others see it as a replacement (Sterling
2004).
Early writings on adaptation emphasized that the most important part of this process is maintaining flexibility to responds to future challenges. Bateson (1963) argued that both acclimatization and genetic adaptation involve a process of restoring somatic flexibility.
31 Slobodkin (1968) argued that adaptation involves a hierarchy of responses where those environmental stressors that cannot be dealt with behaviorally will necessitate a physiologic or developmental response, and those that cannot be dealt with physiologically will result in differential survival and reproduction, and genetic adaptation. In other words, phenotypic plasticity acts as a buffer of genetic change. Further, he argued that a well adapted population is one in which individuals are more likely to survive the environmental conditions they are most likely to encounter. Both of these authors emphasize that adaptation is relative or situation dependent, and that it involves substantial costs which affect an organism’s ability to deal with future events. While most genetic mutation is expected to be neutral or deleterious, it is also the only source of novel, and potentially adaptive traits.
A major difficulty associated with this conception of the relative nature of adaptive processes is defining and measuring adaptation. Mazess (1975) proposed that adaptations be identified based on their positive contribution to one or more of several domains, including reproduction, health, nutrition, the nervous system, work capacity, growth, and reproductive performance. In this way, the definition of a trait as adaptive is entirely dependent on the domain one chooses, and a trait that is adaptive in one domain may be costly in another
(Goodman and Leatherman 1998). Frisancho (1993) proposed distinguishing between adaptation and accommodation, which he defines as a necessary response which increases the likelihood of survival, but which comes at a substantial physiological cost. A common example of this is reduced and delayed growth among children, which may be necessary under conditions of nutrient restriction and infection, but leads to an increased risk of morbidity and mortality
(Frisancho 1993). That biological anthropologists were at the forefront of rejecting the small but health hypothesis (e.g. Martorell 1989; Pelletier 1994), or the idea that small body size is
32 adaptive under conditions of poverty, highlights the selective nature of critiques of research on adaptation in biological anthropology.
Challenges to the concept of adaptation have come from several quarters. In a seminal paper, Gould and Lewontin (1979) argued that the “adaptationist programme” is characterized by a blind faith in the power of natural selection to shape organisms to optimal form and a failure to consider alternatives to adaptive explanations such as pleiotropy and allometry. These authors suggested that traits should not be considered adaptive solely on the basis of biologic plausibility, but that this should be among a number of possible explanations considered.
Kimura (1985) proposed that while adaptation is important at the phenotypic level, most genetic change is the result of drift, and of no selective importance.
West Eberhard (2003) argues that, while genetic mutation is an important source of variation in adaptive evolution, phenotypic novelty can occur by other means, for example, through the effects of the external or internal (other genes) environment on gene expression.
Genetic accommodation is the process by which novel traits expressed in development are subject to the action of selection. This process may involve the regulation of the trait to change its frequency or the conditions of its expression, it may improve the integration or efficiency of the trait, or it may act to reduce the negative side effects of the trait (West Eberhard 2003). This does not imply that traits become subject to greater genetic control. Genetic changes can make the trait more likely, not because the trait is fixed, but because it becomes a common response to specific types of environmental stimuli. In this way, she suggests, environmentally induced novelties, brought about through phenotypic plasticity, lead the way in evolution, and genetic change, in general, follows as the result of genetic accommodation to these new traits (West
33 Eberhard 2003). Developmental plasticity, does not buffer genetic change, but is a fundamental part of the process of adaptation.
West Eberhard (2003) also argues for an altered conception of homeostasis, which shares similarities to the idea of allostasis. Specifically, she suggests that changes in set points in homeostatic mechanisms, brought about through environmental or genetic change, can be an important evolutionary process. This means that under certain conditions phenotypic change may be exaggerated rather than buffered.
Finally, a number of authors have challenged the idea that organisms passively reflect changes in an autonomous environment as, some argue, was suggested by Darwin (Griffiths and
Gray 2001; Lewontin 2000). Organisms, to a large extent, create their own environments by, among other things, selecting those aspects of their surroundings that are relevant, actively altering their surroundings, and creating resources for their own consumption (Lewontin 2000).
A recent focus on the importance of niche construction, has illustrated the dialectical relationship between environmental change and adaptive mechanisms, such that one cannot be considered without the other (Leland and others 2001). Organisms can play an important role in altering their environment in a way that changes the action of natural selection (Leland and others 2001).
This discussion illustrates the complexity of the conception of adaptation within evolutionary biology, and suggests that much of the criticism is oversimplified. Thomas (1998) argues that the traditional focus of human biology on adaptation to environmental stressors, including an increasing focus on psychosocial stress, is still a useful framework, but that we need to expand our notion of “environment” to include larger scale forces, specifically the conditions that create poverty and its downstream effects. The work of biologists on integrating developmental and evolutionary approaches provides an important framework for thinking about
34 the importance of the environment in shaping the phenotype, and for getting past a lot of the false dichotomies (nature/nurture) that obscure our understanding of the natural world (Levins and Lewontin 1985; Lewontin and Levins 2007; Oyama 1985).
Political-economic approaches in human biology
According to Lewontin and Levins (2007: 91), “the art of research is the sensitivity to decide when a useful and necessary simplification has become an obfuscating oversimplification.” No one investigator can possibly expect to have enough resources and expertise to consider the full complexity of human biology, but fortunately, science is a cumulative and collaborative process. Historical political economy has critical insights to offer to the study of human biology, and many of the tools necessary to achieve this synthesis already exist. In this section, I will explore some of the diverse methodological approaches currently employed in biocultural research.
A common approach in biocultural analysis involves using ethnographic techniques such as participant observation and interviews to develop both a qualitative understanding of cultural context, as well as quantitative, individuals level measures of aspects of culture to link to health outcomes (Dressler 2005). Specific examples include the studies of status incongruity discussed above. Some authors have added to this explicit discussion of the political economic factors associated with the relationship between poverty and biology (Crooks 1997; Crooks 1998;
Crooks 1999; Thomas 1998). Leatherman, in particular, has been a consistent advocate of political economic approaches to human biology research, and has made many important contributions to advancing this body of work in the Andes. He has explored the importance of changing labor patterns for seasonal shortages (Leatherman and others 1988), the effects of land reform and socioeconomic change on secular trends in growth (Leatherman and others 1995),
35 and the dialectical relationship between illness, poverty and household production (Leatherman
2005; Leatherman 1996; Leatherman 1998b). Importantly, Leatherman has suggested that a more critical and reflexive biological anthropology can make important contributions to understanding poverty, for example, the relationship between failures in household production and reproduction with the rise of Sindero Luminoso in Peru (Leatherman 1996).
Schell and colleagues have used a partnership approach in conducting research on toxicant exposure and health among the Akwesasne Mowhawk, which provided them with unique tools for modeling biocultural interactions (Schell and others 2005; Schell and others
2007). In order to avoid research methods that provided little benefit to the local community, these investigators worked closely with community organizations at each stage of research, and trained local researchers to carry out much of the data collection. These authors argue that this approach allowed them to collect data in a way that was culturally appropriate, and made use of local expertise in order to design questions to assess relevant aspects of culture. In particular, they are interested in investigating contradictions between the importance of traditional subsistence activities in maintaining a sense of cultural identity, at the same time that they increase the exposure of individuals to industrial toxicants and promote health inequalities
(Schell and others 2005). A further benefit of this approach is its reflexivity and concern for the rights of its participants, demonstrated particularly through their extensive consultation with the community to establish overlapping interests, and through their giving the community a high degree of control over the dissemination of the research findings (Schell and others 2005; Schell and others 2007).
Finally, Dressler and colleagues suggest that while culture clearly has important implications for human biology, explaining variation in objective measures of health can only be
36 accomplished by connecting the biological and the cultural at the level of the individual (Dressler and Bindon 2000), in a way that does not conflate culture with the beliefs and values of individuals (Dressler 2005). In order to accomplish this goal, these authors draw on theory and methods developed in cognitive anthropology, which allow the development of cultural models by determining the degree of overlap between individuals in their personal ideas about how society works (Dressler 2005). Cultural consensus analysis measures the degree to which people are drawing on a shared cultural model of a particular domain, for example, which sources of social support are appropriate in a given situation (Dressler and others 1997). Cultural consonance measures the degree to which individuals are able to approximate this shared cultural model in their behavior (Dressler and Bindon 2000). Dressler and colleagues have demonstrated the importance of cultural consonance in both social support and material consumption for blood pressure in diverse populations (Dressler and others 1997; Dressler and Bindon 2000).
Particularly important is the use of consensus methods to explore the influence of the cultural construction of race on health (Dressler and Bindon 2000; Gravlee and Dressler 2005; Gravlee and others 2005).
Dressler makes an important point that issues of measurement are not a trivial consideration in biocultural anthropology, and has done much to advance systematic and sophisticated means to quantify culture and relate it to health (Dressler 2005). Though he does not explicitly argue against the incorporation of political economy into human biology, Dressler
(2001: 457) has advocated against the replacement of current models with a “pseudosynthesis that simply involves the layering of potentially relevant influences on health into ever larger models.” His critique is well taken but I argue, underestimates the theoretical importance of historical political economy for understanding health inequalities. Certainly measurement issues
37 will have to be addressed, but this will only happen if we value this perspective enough to invest in its development. Further, the value of a qualitative understanding of these larger forces cannot be overestimated (Lewontin and Levins 2007). A theoretical grounding in regional and global political economy is critical in determining what questions we ask, how we ask them, and how we interpret our data.
With regard to methods, it is clear from the preceding discussion that there are a number of different approaches to biocultural research that have provided important insights into human biological variation. What I would add to this general discussion is that we need to continue to make use of the methods that anthropologists have long employed in understanding culture, including intensive ethnography over long periods of time. A sophisticated understanding of local conditions often comes with long term investment, coupled with detailed regional knowledge. In terms of linking political economy with health; we need to develop theory that allows specific predictions about the links between culture and political economy, and their dialectical relationship with biology. It is important not to think of these large scale forces as operating separately from one another or somehow set apart from everyday life, but as fundamentally interpenetrating, and critical for understanding conditions at the local level. I believe that with a systematic effort, the methods outlined above can be adapted for this purpose.
My aim in this dissertation is not to resolve these issues, but rather to begin to address them.
More specifically, carrying out my dissertation research and analysis has convinced me of the importance of political economy. Throughout this dissertation I will try to address the goal of linking large and small scale influences on health to the extent possible with these data, but more importantly, I will discuss what this work has taught me about how to approach this goal in the future.
38 El Alto, Bolivia, the city in which I conducted my dissertation research, is today widely known as a center of leftist political activism, and a critical base of support of the country’s first indigenous president, Evo Morales. Generally speaking, people in El Alto are well aware of the effects of neoliberalism in their everyday lives, and are concerned about inequality on many fronts, just one of which is health and access to healthcare. I believe that a political economic perspective on human biology is a critical component in conducting research that is relevant to the people of Bolivia. It was with this intention that I investigated the importance of social support for women’s health in El Alto. In the following section I will discuss some of the specific structural concerns for people in Bolivia which promote economic, and likely, health disparities.
Neoliberalism in Bolivia
39
Figure 1.1. Relief map of Bolivia.
El Alto is often described as a marginal city, but this is an economic distinction rather than a political one. In the last decade, this urban center has drawn an enormous amount of international attention as the site of protests which helped force the resignation of two presidents in as many years, and as an important base of support for the country’s populist, indigenous president Evo Morales. Part of the “pink tide” of left leaning political shift in Latin America, social movements in Bolivia are well aware of the unequal terms of free trade which favor a small number of local elites, foreign governments, and corporations over the country’s majority indigenous population, and have mobilized to contest this global order. Following Roseberry’s
(1982) argument that culture is socially constituted and socially constitutive, I will argue that
40 cultural norms of social support and social relations in El Alto cannot be divorced from an understanding of the forces that serve to constrain household production and economic mobility, and the cultural means through which these forces are contested. That is, cultural expression is both about establishing narratives which serve to support or enhance the status quo with regard to economic and social relations, as well as narratives which challenge this status quo. In
Bolivia, a sense of imagined community (Anderson 1983) built around idealized notions of indigenous history and identity, among other things has served to bring together diverse groups for the purposes of protesting and resisting neoliberal hegemony (Albro 2006); though social movements in Bolivia are associated with a number of contradictions, they have been relatively successful in making their voices heard.
Despite a wealth of natural resources, Bolivia is one of the poorest countries in Latin
America, stemming in part from a long history of brutal colonial rule (Dangl 2007). The Spanish empire was financed first on silver, then on tin mined by workers in virtual slavery, and owners of haciendas accumulated wealth through their dependence on the forced labor of indigenous people, resettled without respect to kinship or ethnicity (Klein 1982; Larson 1998). It was only during the early 1950s, and after a revolution bringing together the urban middle class and rural campesinos , that indigenous people won the right to vote, and gained property rights to the land they worked (Klein 1982). Over several decades, military dictators provided wealth and resources to political supporters and ran up the national debt. Paradoxically, democratic elections in 1982 provided a greater voice to citizens of Bolivia, but also ushered in the start of a program of neoliberal reform that was the most drastic on the continent, with profound consequences for the economic rights of the same people who had only recently become full citizens (Gledhill 2006).
41 Neoliberal economic theory holds that economic growth and public well being will best be served by minimal government intervention in the economy, and the free reign of market forces. The proper role of the state in this vision is to protect rights to private property and entrepreneurial freedom, through violence and coercion, if necessary, without regard for popular will (di Leonardo 2008; Harris 2002; Harvey 1990; Harvey 2005; Harvey 2006; Klein 2007).
Harvey (2005) traces the foundations of neoliberalism back to the presidency of Jimmy Carter in the late seventies, and through the tenures of Ronald Reagan and Margaret Thatcher, but argues that the neoliberal orthodoxy known as the Washington Consensus was only reached in the
1990s, under the watch of president Bill Clinton and prime minister Tony Blair, members of the
Democratic and Labour Parties, respectively. Today, the idea that economic development in the global south depends on deregulating markets to attract foreign investment and removing all barriers to the flow of capital, is largely unquestioned within the development community and the international bodies which concern themselves with these issues (Harvey 2005; Kohl and
Farthing 2006). A critical component of this orthodoxy, that unrestricted capitalism is the most efficient way to reduce poverty and improve living standards around the globe, is being met with increasing skepticism among the majority of people in countries like Bolivia, who have seen their incomes continue to decline over the last twenty years (di Leonardo 2008; Gledhill 2006).
After the stock market crash of the 1930s, which demonstrated the inherent instability of unfettered capitalism, and the rapid economic growth experienced in the US during World War
II, western nations adopted Keynesian economic policy in the interest of maintaining economic stability and preserving public welfare. Under this model, states would intervene in the economy to ensure full employment and economic growth, and to minimize the effects of economic downturns (Harvey 2005). This period was characterized by the development of substantial
42 regulation of business, and in some countries, direct participation of the government in the market in the form of state owned and operated corporations (Harvey 2005). Often referred to as
Fordism, there was general agreement that economic growth would be sustained through the purchasing power of a substantial middle class, and that workers would earn wages and benefits sufficient for social and economic mobility (di Leonardo 2008; Harvey 1990), at least in part, to prevent citizens from being attracted to radical left, or rightwing social movements as happened in Germany after the First World War (Klein 2007).
As economic growth began to slow in the seventies, the debate surrounding the limits of government involvement in the economy, including government oversight of corporations, was rekindled (Harvey 2005). Keynes had advocated, in particular, increasing government spending during times of economic downturn, and paying down government debt during times of prosperity. Neoliberal theorists led, in particular, by Milton Friedman and colleagues at the
University of Chicago, argued against Keynsian economic theory, suggesting that while governments could never understand all of the variables well enough to intervene effectively, the market could be trusted to regulate itself to the benefit of all (Harvey 2005; Klein 2007). Indeed,
Friedman, and like minded economists, believed that a capitalist system free of all government interference would reach an equilibrium such that prices and wages would balance each other perfectly, employment would be high, and inflation non existent (Klein 2007). While their mathematical models supported this prediction, observing a real world application of radical neoliberal reform became a top priority for these academics (Klein 2007). American government priorities began to converge with those of this group as increasing economic nationalism in the developing world raised the threat of appropriation of assets from US corporations (Harvey
2005).
43 Some regions of the global south experienced little change with neoliberal reform, as these countries had limited economic power for Keynesian interventionism to begin with, however; this was not the case in South America (Klein 2007). Countries such as Argentina and
Chile pursued protectionist economic policy along with investment of public money in developing business and infrastructure, practices which led to considerable economic growth, high levels of employment, and improvements in living standards for a large proportion of these populations (Klein 2007). This was true to a lesser extent in Bolivia, where, between successive dictatorships, President Paz Estenssoro enacted land reform, introduced universal suffrage, and nationalized the country’s tin mines (Klein 1982; Klein 2003). In a unique collaboration designed to counter these trends in Chile, USAID provided funds to bring Chileans to study at the University of Chicago, and helped found a department of economics at The Catholic
University in Santiago (Klein 2007). When neoliberal ideas failed to gain popularity within the context of the democratic process, members of the CIA provided training and support to General
Augusto Pinochet to stage a violent and bloody coup in 1973. Over the next decades, overseen by Chicago trained economists, Chile served as a laboratory for radical privatization, deregulation and cuts in social spending (di Leonardo 2008; Harvey 2005). Despite the massive damage done to the economy, including increases in rates of poverty, unemployment, and the national debt, these policies were successful in terms of transferring wealth from the middle and working classes to a small group of local beneficiaries and multinational corporations (Harvey
2005). As such, Chile served as a model both for other developing countries, in what would come to be known as Structural Adjustment Programs (SAP), and the adoption of these measures became a condition of aid from western nations, The World Bank, and the International
Monetary Fund (IMF) (Harvey 2005; Kohl and Farthing 2006).
44 Shortly after a sustained period of military rule, the Bolivian economy in 1985 was characterized by an enormous national debt and skyrocketing inflation (Klein 2003). A close election in this year brought to power Víctor Paz Estenssoro, a fourth term president and well known economic nationalist (Klein 2003). Though little was known at the time about how he developed his economic policy, recent reports from participants indicate that Paz convened a bi partisan committee to address the economic crisis, a fact which he kept secret even from his own cabinet (Klein 2007). Working based on the policy prescriptions of Harvard economist Jeffrey
Sachs, and under the direction of Gonzalo Sanchez de Lozada, trained at The University of
Chicago, this committee hammered out the Supreme Decree 21060, or the “New Economic
Policy.” This policy was met with great satisfaction by members of the US government and the
IMF, ensuring uninterrupted flow of foreign aid (Gill 2000; Klein 2007; Kohl and Farthing
2006). What made their approach different from those seen in countries like Chile, was the sweeping and sudden nature of the reform: in this single decree were 220 separate laws dealing with, among other things, devaluation of the currency, the termination of price and wage controls, the elimination of public sector employment, and a massive reduction in government spending including in areas such as health care and education (Gill 2000; Klein 2003).
This New Economic Policy, described by Gill (2000: 12) as “the opening salvo in an assault on the poor,” was successful in the sense that it stabilized the economy and put an end to the soaring inflation, but it also reduced economic growth and dramatically increased levels of poverty (Healy and Paulson 2000; Klein 2003). Neoliberal economists, including Sachs, theorized that rapid and dramatic changes, like pulling off a bandage, would lead to quicker economic recovery, including growth in private sector employment for former government employees, and a quick stabilization of wages with respect to basic commodity prices (Klein
45 2007). These benefits, however; never materialized. Even decades later, wages, benefits and job security are low while unemployment remains high, and there are very few state programs to mitigate poverty. Particularly hard hit were the miners fired during the closing of government owned tin mines and forced to migrate in search of alternative employment. Prior to this dispersal, miners had been extremely well organized and aggressive opponents of the state, and had performed an important leadership role for labor in the country (Gill 2000; Nash 1994).
Bolivia was held up as a success story by the international development community because this was the first administration to achieve such massive changes in such a short period of time, and, more importantly, Paz was able to accomplish this program within a democratic system (Klein 2007). This conclusion is highly suspect, however; as there was no mandate or public consultation for this change of direction, and its implementation required a substantial amount of government repression, including the imprisonment of labor leaders in the eastern lowlands until all the measures had been implemented (Gustafson 2002; Klein 2007). There was an enormous amount of public protest of these measures, but as in many other nations, Bolivians continued to elect neoliberal governments over the next several elections (Healy and Paulson
2000).
A slightly left of center party was elected in 1989, but failed to challenge the economic model set out by Paz. The reelection of the MNR, headed this time by Gonzalo Sanchez de
Lozada (A.K.A. Goni), in 1993 ushered in another period of neoliberal reform which would have profound effects on a number of fronts including the nature of democratic participation and citizenship (Healy and Paulson 2000; Kohl 2003). In “ El Plan de Todos ” a large number of changes were enacted, including constitutional reform, judicial reform, and the introduction of a social insurance program for individuals aged 65 or older, but perhaps the most important
46 changes came with the Law of Capitalization and Law of Popular Participation, both passed in
1994 (Kohl 2003; Kohl and Farthing 2006).
The Law of Capitalization, designed to create jobs and encourage foreign investment, involved the partial sale (50%) of the stock in state controlled industries including oil and gas, telecommunications, airlines, electricity, and railroads. Together, these companies provided
60% of national revenues, ensuring that the government would have to rely, to a large degree, on foreign aid for any social spending or public investment, and would thus be more vulnerable to international pressure (Kohl 2002; Kohl and Farthing 2006). This law did encourage foreign investment, but without providing employment to Bolivia’s workers. Multinational corporations, focused on the bottom line, improved their profits by firing workers or reducing wages, importing necessary manufactured products rather than having them made in Bolivia, and providing limited financial compensation to the local community for the extraction of natural resources (Kohl and Farthing 2006). At the same time, newly privatized companies charged more for their services than had the government, contributing to rapid elevations in the cost of living.
There are two factors which kept the economic effects of these two periods of reform from being as disastrous as they might have been early on: many displaced workers migrated to the Chapare , or the forested highlands of the eastern Andes, to farm coca, whether for domestic consumption or for the international cocaine market; others migrated to Argentina to work, sending remittances back to family in Bolivia. As part of the war on drugs, the US has long made foreign aid dependent on aggressive coca eradication projects. Starting in 1998, the former general Hugo Banzer launched a highly militarized program, called Coca Zero, designed to eradicate all but a small portion of coca crops for domestic use (Kohl and Farthing 2006). In
47 1999, the collapse of the Argentine economy had a profound effect on the flow of money into
Bolivia, further eroding alternative coping mechanisms, and worsening the economic situation in the country (Kohl and Farthing 2006).
The Law of Popular Participation involved the transfer of 20% of the national budget to municipalities, which then became responsible for spending on health, education, infrastructure, micro irrigation, and sports facilities. Neighborhood and Indigenous organizations became legally recognized representatives of their communities as Grassroots Territorial Organizations
(GTOs), and now had the opportunity to make proposals regarding how the municipal budget would be spent. In addition, an oversight committee, comprised of members of GTOs now had the ability to veto municipal budgets. In some areas, strong grassroots movements have been able to use this system to their advantage, but in others, GTOs are not being given the opportunity to plan or approve budgets by local politicians (Kohl 2002). In addition, the size of the municipal budgets don’t reflect the large number of responsibilities newly transferred to municipalities (In 1999, municipalities received 26$ per capita), and the law includes relatively strict rules regarding which types of programs can be financed and what percentage of the budget can be used for each (Kohl 2002). This law is, in many ways, consistent with neoliberal principles regarding the decentralization of democratic processes; it also helped to enhance grassroots organizing around ethnicity and indigeneity, among other things (Albro 2006; Kohl and Farthing 2006).
There are numerous contradictions associated with neoliberal policy, in practice and in theory (Harvey 2005). The first is that, while much of the discussion is focused on reducing the role of government in the economy, in practical terms, it is not so much reduced as it is re directed toward helping those at the top end of the income distribution, often described as
48 “corporate welfare” (Harris 2002). This is true at the national level, but also at the international level, as western nations continue to intervene in their economies to protect whole industries, while exerting influence on lesser developed nations to completely abandon these practices. The second, and related to the first, is that though neoliberal theorists focus on the absolute sanctity of individual freedom of consumers to pursue their economic self interest, individual freedom is often subverted to that of capital, particularly with regard to ownership and private property
(Harvey 2005; Klein 2007). The third is the idea, popular in the development community including the IMF, World Bank and USAID, that capitalism and democracy are inherently compatible, and even mutually reinforcing. The successful development of a capitalist economy under a totalitarian regime in China is only one example of falsity of this connection (Klein
2007). Indeed, Klein (2007) argues convincingly that neoliberal reform packages are rarely supported by democratic majorities, but have to be implemented during periods of shock. In
Bolivia, Kohl and Farthing (2006) argue that a specific kind of democracy, that is, one focused at the municipal rather than regional level, reduces government power to interfere in the economy, and helps direct protest away from national and regional governments. Others have focused on the potential of human rights and multicultural discourse, centered on identity politics, for weakening or strengthening solidarity between those most affected by these policies (Gledhill
2005; Gustafson 2002; Healy and Paulson 2000). Finally, fourth is the pressure on developing nations to adopt these measures to maintain the flow of foreign aid, however; this sometimes disguises the fact that net capital flows from developing to developed nations as it did during the colonial period (Nash 1994).
In her film Our Brand is Crisis , Rachel Boynton (2006) captures the contradictory nature of international involvement in Bolivian politics and economy. This documentary follows the
49 work of Greenberg, Carville, and Shrum, a consulting firm responsible for the elections of Bill
Clinton and Nelson Mandela, in their effort to help re elect Goni to office in 2003. In discussing their preferred candidates, Jeremy Rosner describes their alliance with politicians who are progressive and social democratic, and who emphasize market driven policy, but with benefits for the entire society. Throughout the film, he conveys his genuine belief that Goni is the best candidate to help Bolivia escape economic crisis, and the real danger of the rhetoric from then presidential candidate Evo Morales, making promises on which he would be unable to deliver.
With a heavy reliance on focus groups to test messaging, and negative campaigning against their closest opponents, this firm helped elect Goni with 22.5 percent of the national vote (Evo received 20.9%, Manfred 20.8%). Tellingly, when asked by a reporter whether he would consult with the public on plans to export gas to the US through ports in Chile, Goni stated “no, you can’t ask the people about such complex issues. That would be an injustice, like using a blunt knife for surgery” (Boynton 2006). It was, of course, his refusal to listen to the overwhelming majority of Bolivians who opposed this policy, which caused massive protests, dozens of deaths from police violence, and his eventually having to flee the country.
Despite the overwhelming evidence that neoliberal economic policy dramatically increases inequality (di Leonardo 2008), leaving the majority worse off, neoliberalism is widely accepted as common sense, especially by those providing resources and support for development projects (Harvey 2005; Kohl and Farthing 2006). This explains how a politician focused on redistributing wealth upward, without reference to public opinion, can be described as a progressive social democrat. It also explains how workers with USAID and other agencies can suggest that economic problems in Bolivia are about a lack of democracy at the same time that they reject the legitimacy of the majority of political candidates. That is, the accepted link
50 between free market economics and democracy helps make capitalism more palatable (Kohl and
Farthing 2006), but economic policies which benefit western corporations always takes precedence over the democratic process.
There are similar contradictions involved in the operation of NGOs in developing nations. Gill (1997; 2000) argues that there were a small number of progressive NGOs in operation in Bolivia during military rule, mainly focused on achieving democratic reform, but starting with democratization, they began to work in closer alliance with the state and with funding organizations, a process which considerably limits their ability to be critical, and politically engaged (Gill 1997; Gledhill 2006). While the provision of basic services and economic aid assists local communities in the absence of state services, NGOs provide limited benefits, focus preferentially on some sectors of the population and on some localities, leaving many to do without (Gill 1997). These organizations also take pressure and focus off of the state as a service provider as they cut social spending (Gill 1997; Gill 2000; Gledhill 2006; Kohl and
Farthing 2006). NGOs have expanded as an alternative form of employment for a professional middle class, following cuts in public sector employment, leaving many in cities like El Alto under the impression that these agencies serve as a means of obtaining grants primarily for salaries, rather than for community services (Gill 1997). Finally, these agencies tend to focus on micro enterprise or micro credit operations, often by providing high interest loans and training
(Gill 1997), however; economies of scale for larger corporations ensure that these small businesses will be less competitive in a free market (Kohl and Farthing 2006). There is nothing inherently bad about NGOs and many employees for these agencies feel a deep sense of responsibility toward their clients, but their embeddedness within the neoliberal system helps to perpetuate this system (Gill 1997; Gledhill 2006; Kohl and Farthing 2006).
51 While neoliberalism represents a hegemonic discourse in much of the globe, it is also highly contested; and nowhere more so than in Bolivia (Gledhill 2006; Kohl and Farthing 2006).
Starting in 2000, with the water wars in Cochabamba, coalitions of indigenous, labor, and campesino organizations have come together to successfully protest the neoliberal turn of the country, reversing the commoditization of water in Cochabamba and El Alto; resisting tax hikes; preventing the export of natural gas through Chile; and; electing the country’s first indigenous president, running on a platform of nationalizing the oil and gas industry, land redistribution, and greater autonomy and civil rights protection for indigenous communities (Dangl 2007). Much of this protest has taken place within El Alto. A strategic site overlooking the seat of government in
La Paz, protesters both from inside and outside of El Alto converged here to blockade the city and force government recognition of their anger and distress over the failures of neoliberal economic policy (Dangl 2007).
Despite the success of indigenous and other social movements centered in the western highlands, in terms of arresting privatization and winning political power, elites, particularly in the eastern lowlands, have fought back against Morales’ economic nationalism by campaigning for regional autonomy, and greater control over local resources. This drive toward regional autonomy threatens many of Morales’ most important goals, including increasing state revenues from the oil and gas sector, and the redistribution of unproductive land in the eastern lowlands to small scale farmers. In this context, cultural production has important material effects in terms of helping determine who gains access to the country’s natural resources, including whether more of the revenues go to Bolivians instead of multinational corporations, and whether resources are more equitably shared amongst the country’s residents. The potential for decreasing political and economic inequality has important implications in terms of health
52 inequalities. In this dissertation, I not explicitly examine health inequalities on the basis of racial or economic inequalities, but rather the influence of macro economic conditions on social support, and the importance of social support given the retreat of the state in Bolivia.
Political economy and women’s health in El Alto
As in other parts of the global south, economic difficulties in rural areas of Bolivia have forced many of Bolivia’s indigenous people to migrate to urban areas such as El Alto. However, the harsh realities of life in this peri urban community include limited access to formal employment and inadequate social services (Arbona and Kohl 2004; Gill 2000). Women are particularly vulnerable as they must balance economic activities with childcare, and are increasingly likely to head their own household (Feldman and others 1992; Miller 1995). At the same time that indigenous people struggle to overcome economic injustice imposed by the federal government in collaboration with foreign governments and institutions, they contest cultural narratives which place the blame for economic inequality on the victims of this inequality, and struggle to use what resources they have to make claims regarding rights of citizenship and access to economic opportunity (Goldstein 2004; Lazar 2008). In addition to political struggle and involvement in the informal economy, I suggest this may include social solidarity in an economic situation that tends to reinforce household retrenchment.
Reciprocal relationships based on kinship and fictive kinship are an important part of economic life in the rural Andes, and constitute a form of wealth developed over the course of a lifetime of rural production (Weismantel 1988). Collins (1986) has argued that involvement in the wage economy weakens these extended networks, at the same time that inadequate wage earnings necessitate a diversified economic strategy, and continued reliance on social ties. While relatively less is known about social networks in urban areas of the Andes, ethnographic work
53 suggests that social ties may be an important resource for individuals trying to establish themselves in El Alto (Buechler and Buechler 1971). I hypothesize that social support, particularly instrumental social support, is a centrally important coping mechanism in this marginal urban setting, and that social support will have implications for understanding variation in women’s health. My dissertation research set out to address the following objectives:
1) To examine the relationship between household composition, poverty, and health among women in El Alto, Bolivia.
The feminization of poverty in urban areas of the developing world is related in part to the increasing number of female headed households (Feldman and others 1992). Few projects have investigated whether the increased risk of poverty associated with single parent households is associated with poorer health outcomes for women. This dissertation will compare the health of women from male, female and dual headed households.
2) To examine how culturally relevant social support influences multiple dimensions of women’s health.
While social support has long been an important area of research in public health, the cultural dimensions of social support are still poorly understood (Dressler and others 1997; Dressler and
Bindon 1997; Jacobson 1987; Janes and Pawson 1986). I used participant observation and semi structured interviews to develop a structured questionnaire to examine variation in sources and types of support for women in El Alto, particularly focusing on forms of emotional and instrumental support. This dissertation will examine the potential influence of social support on nutritional status, immune function, and inflammation in this setting.
3) To examine how household composition and socioeconomic status influence the relationship between social support and health.
54 Finally, given the relative poverty of women who head their own households, and the high rates of poverty in El Alto, this dissertation will explore possible interactions between social support and either household composition or socioeconomic status in predicting health.
Specifically, I will examine whether social support is a more important predictor of health for single mothers, than for women living in other circumstances, or for women in more economically marginal households.
In the next two chapters I will explore the history of human biology in the Andes, focusing on political economic approaches; recent work on urbanization and health; the biological pathways through which social support may influence health, and; cultural norms among Andean populations which may be relevant for social support in El Alto. I will then move on to examine the objectives outlined above.
55 Chapter 2: Urbanization and human biology in the Andes
Introduction
Changing theoretical perspectives in the study of human biology in the Andes reflect broader currents within the discipline. Early research was primarily focused on understanding the physiological and genetic responses of humans to the unavoidable stressor of reduced oxygen pressure (Beall 2001). While much of the research produced in the Andes continues in this vein, starting in the 1960s, a number of investigators began to consider the importance of multiple sources of stress for human biological variation, the most important of which appears to be poverty (Leatherman 1998a). Thomas and colleagues (1988), working in the Peruvian Andes, were among the first to take on the challenge of integrating bio cultural with political economic perspectives, and their project provided important insights into the implications of changing labor and social relationships under capitalist development, for health. I will begin this chapter by reviewing human biology in the Andes, with specific reference to political economic approaches.
The same forces that lead to the marginalization of rural producers, influence the incredible rate of migration to urban areas, while simultaneously limiting urban livelihoods (Gill
2000). As reflected in a volume edited by Schell and colleagues (1993b), human biologists have long been interested in understanding the relationship between urbanization and health, but we have only begun to think about how to recognize the complexity of urbanization, as the localization of global processes (Roseberry 1982). I will discuss studies of urbanization and human variation within biological anthropology and public health, and how a political economy of human biology can make critical contributions to this body of research. Finally, I will discuss
56 the implications of changing social norms associated with urbanization, and how these might influence the relationship between social support and health.
Human biology research in the Andes
High altitude hypoxia is caused by decreasing barometric pressure at higher elevations.
Oxygen diffuses from the alveoli of the lungs to the blood stream, and is carried by hemoglobin to the rest of the body. Under normal circumstances, this diffusion is promoted by the higher partial pressure of oxygen in the lungs, while the lower partial pressure in the tissue capillaries promotes the release of the oxygen molecules into the tissue. Cellular respiration, the processes by which oxygen is consumed by mitochondria to create ATP, the molecule responsible for cellular work, requires a constant supply of oxygen. At 4000 meters there is a 37% reduction in the number of oxygen molecules in a liter of inspired air when compared to sea level (Beall
2001). As a result, there is a reduced amount of oxygen available to diffuse into the blood and oxygen saturation of hemoglobin is reduced.
The first step in the acclimatization of sea level natives to high altitude involves increased pulmonary ventilation and heart rate, stimulated by falling levels of blood oxygen (Frisancho
1993). Over the course of days, the release of erythropoietin stimulates increased production of red blood cells and hemoglobin, and increases blood volume, a response associated with an improvement of the oxygen carrying capacity of blood (Frisancho 1993). Low altitude natives encountering high altitude for the first time are at greater risk of health complications such as acute mountain sickness, or high altitude pulmonary edema, which if untreated, can result in death (Frisancho 1993). While acclimatization is possible, low altitude natives cannot achieve the level of function seen in permanent residents of high altitude, and demonstrate poorer
57 physiological functioning, lower work capacity and poorer birth outcomes (Beall 2001;
Frisancho 1993; Moore and others 1998).
Recent work has exposed some additional responses of low altitude natives to hypoxia.
Inflammation is increased at high altitude as evidenced by rises in interleukin 6, interleukin 1, and C reactive protein (Hartmann and others 2000). This inflammation likely leads to decreased fluid clearance in the lungs, which may play an important role in the development of high altitude pulmonary edema (Basnyat and others 2001; Hartmann and others 2000). Pregnant women at living at high altitude in the US have increased levels of catecholamines and pro inflammatory cytokines compared to non pregnant women, which are implicated in the development of preeclampsia and premature delivery (Coussons Read and others 2002;
Coussons Read and others 2005). T lymphocyte function is mildly impaired at high altitude, a response which likely limits the body’s ability to fight infection (Basnyat and others 2001).
Finally, work on mountaineers has demonstrated that hypoxia is associated with a significant loss of muscle, as well as a breakdown and decrease in density of mitochondria (Hoppeler and others
2003). This is probably due to the damaging effects of reactive oxygen species, but the physiological mechanisms associated with this process are still poorly understood (Gelfi and others 2004; Hoppeler and others 2003).
Andeans, Tibetans, and Ethiopians demonstrate different suites of adaptation to high altitude, suggesting multiple solutions to similar sources of environmental stress (Beall 2001;
Beall and others 2002). Among Andean populations, hemoglobin concentrations are elevated relative to populations at sea level, particularly among males and those living in urban areas
(Beall and others 1998; Tarazona Santos and others 2000; Vasquez and Villena 2001). This response is associated with an increased risk of polycythemia, or excessive blood viscosity due
58 to abnormally high hemoglobin levels (Beall and others 1998). Oxygen saturation, or the percentage of hemoglobin carrying oxygen, ranges from 95 % to 99 % among sea level natives
(Beall 2001). Beall and colleagues (1999), found that Andeans have 2.6 % higher oxygen saturation than Tibetans living at similar altitudes, but both had lower levels than seen in low altitude populations. High altitude natives from the Andes demonstrate better oxygen saturation during exercise than Europeans born and raised at high altitude, possibly due to larger lung volume and improved oxygen transport in the lungs (Brutsaert 2001; Brutsaert and others 1999).
Finally, Andeans tend to have pulmonary artery hypertension due to vasoconstriction and tissue reinforcement of pulmonary arteries, a response which appears to be highly reversible upon migration to lower altitudes (Beall 2001). Beall (2001) argues that the function of this trait is to increase pulmonary diffusion of oxygen to the blood. An associated cost of pulmonary hypertension appears to be an increased risk of pulmonary edema (Beall 2001).
Infants born at high altitude are smaller compared to those born at lower elevations
(Moore 2003; Moore and others 2004). Europeans and Han Chinese born and raised at high altitude have lower birth weights when compared to Andeans and Tibetans, however; low birth weight and associated complications are more common for Andean than Tibetan women (Moore
1990; Moore 2003; Moore and others 2004). This reduction in birth weight appears to be associated with intrauterine growth restriction rather than pre term delivery (Moore and others
2004). Some authors have suggested that the effects of hypoxia are an important determinant of the lower average birth weights found in Bolivia compared to other countries in South America
(Keyes and others 2003). Intrauterine growth restriction, preeclampsia, and gestational hypertension are important predictors of low birth weight and infant death (Keyes and others
2003). Women in both Colorado and in the Andes increase ventilation during pregnancy, a
59 process which increases oxygen saturation and compensates for a fall in hemoglobin levels with pregnancy (Moore 1990).
The physiological responses of both high altitude and migrant populations to hypoxia, and the costs associated with these responses, suggest that this is an important stressor to which individuals and populations must adapt in some way. However; the early expectations regarding the nature of these adaptations have been somewhat contradicted. Specifically, there is less evidence for genetic adaptation than initially expected, and substantial between and within population variation (Frisancho 1993; Greksa and Beall 1989). In addition, a number of authors have suggested there is too great a tendency to assume that any physiological response is the result of hypoxia. Recent work suggests that concerns over genetic determinism in response to early expectations have led to a general underestimation of genetic contributions to some adaptations to high altitude hypoxia including increased lung volume, suggesting a need to consider more fully different levels of adaptive response (Greksa and Beall 1989). In the following discussion, I will outline the development, and changing focus of human biology research in the Andes.
Greksa and Beall (1989) suggest that the history of high altitude research in the Andes can be divided into three periods: 1) pre Nuñoa (1928 1964); Nuñoa (1964 1978); and post
Nuñoa (1978+), each with its own characteristic focus. Studies during the earliest period emphasized the search for physical, and presumably genetic, adaptations to hypoxia of high altitude natives (Greksa and Beall 1989). These investigators were interested in understanding the superior functioning of Peruvian Quechua relative to low altitude migrants, and assumed that hypoxia was the main stressor to which Andeans are exposed (Hurtado 1932; Monge 1948).
Many of the findings of this period, including the increased chest size and vital capacity of
60 highland Quechua relative to similarly sized lowlanders are still considered important aspects of the adaptation of some populations to hypoxia (Greksa and Beall 1989).
The Nuñoa phase of high altitude research was built on the development of the New
Physical Anthropology. Through much of its history, biological anthropology had been based almost exclusively on human classification and the study of race (Brace 2005; Thomas 1998).
The New Physical Anthropology, which grew out of this tradition, was characterized by the adoption of the Modern Synthesis of Mendelian genetics with Darwinian evolution by natural selection, as well as an increased focus on the influence of ecology on human biological characteristics (Hunt 1981). The International Biological Program (IBP) was established by the
International Council of Scientific Unions and the International Union of Biological Sciences, in order to stimulate international investigation of those problems thought to be most relevant for shaping biology. One important aspect of the Human Adaptability Project (HAP) of the IBP was the investigation of human adaptation to high altitude ecology (Harrison 1966; Lasker 1969).
This was related primarily to the fact that hypoxia was seen as a stressor to which behavioral adaptations would be effectively useless (Beall 2001).
The Nuñoa project, initiated by Paul Baker (1969) in the early 1960, provided research and training opportunities for an entire generation of human biologists. The major findings from this work were summarized in the edited volume by Baker and Little (1976), Man in the Andes.
This work emphasized the multi stress nature of the high altitude environment, including that from cold, solar radiation, undernutrition, disease and acculturation. These researchers were interested in the connection between different physiological systems, the balance between competing selective pressures, and the full range of adaptive mechanisms to these stressors, including behavioral and developmental (Greksa and Beall 1989). In particular, the work of
61 Frisancho emphasized the developmental nature of body proportions and improved pulmonary function at high altitude (Frisancho 1969; Frisancho 1993; Frisancho and Baker 1970; Frisancho and Greksa 1989; Frisancho and others 1973; Greksa and Beall 1989). Greksa and Beall (1989) suggest that contemporary trends in biology, including an increasing concern about genetic determinism, were important factors influencing the speed with which Frisancho’s focus on developmental adaptations were accepted.
An important realization associated with the recognition of multiple stressors, was that confounding is an important problem in investigations of responses to hypoxia (Beall and others
1998; Greksa and Beall 1989). For example, early studies of hemoglobin levels among high altitude natives did not account for the potential confounding of dietary iron (Beall 2001). In addition, researchers began to focus more explicitly on potential behavior adaptation to high altitude settings. Following the work of cultural anthropologists, a number of these authors hypothesized a pattern of Andean behavioral adaptation including a reliance on crops and animals efficient at high altitude, trading of resources between higher and lower elevations, and ecological zones to improve dietary diversity, and social organization that emphasizes reciprocity
(Leatherman 1998a).
Investigators during this period outlined a pattern of growth characteristic of Andean populations including a delayed and poorly defined adolescent growth spurt, prolonged growth into early adulthood, and short adult stature (Baker 1969; Frisancho and Baker 1970). Though they considered the potential contribution of undernutrition to this pattern, based on animal models and some limited data on dietary intake, these authors suggested that hypoxia was the primary determinant (Baker 1969; Frisancho and Baker 1970). Frisancho (1976) argued that, contrary to this general pattern, the increased chest size and lung volumes of Peruvian Quechua
62 are a result of relatively greater energy investment in the growth of the cardiorespiratory system as an adaptation to hypoxia.
The post Nuñoa period has been characterized by a great diversity of approaches, but there are a few important trends. A number of researchers have maintained a focus on understanding the dynamics of adaptation to high altitude hypoxia by comparing the growth and development of different populations of high altitude natives and migrants (Greksa and others
1984; Greksa and others 1988; Stinson 1980; Stinson 1982), and by examining different patterns of adaptation among populations from different high altitude settings, specifically the Andes,
Tibet, and Ethiopia (Beall 2001; Beall and others 1998; Beall and others 2002; Moore 2000;
Moore 2003; Moore and others 1998). These studies have provided insights into the importance of high altitude adaptation in influencing physiological function, work capacity, and reproductive outcomes, and have demonstrated a remarkable variability in adaptive mechanisms both within and across regions. At the same time, in response to political economic perspectives within cultural anthropology, and a growing critique of the “Adaptationist Programme” (Gould and
Lewontin 1979) a number of researchers began to more thoroughly investigate human biology in relation to stressors other than hypoxia, using an explicitly biocultural framework (Leatherman
1998a). I will focus on this last body of work as a foundation for more integrative research in the
Andes.
R Brooke Thomas directed a project aimed at integrating political economic and ecological perspectives on health among rural producers and people living in a small town in
Nuñoa, Peru (Leatherman 1998a; Thomas 1998; Thomas and others 1988). As discussed above, a considerable body of research has been devoted to the stressful nature of the Andean environment, however; Leatherman (2005) has argued that the success of past Andean
63 civilizations suggests that political economic constraints are a critical determinant of people’s ability to cope with these ecological challenges. Of particular interest to these researchers was investigating dialectical relationships between poverty, subjective illness, and social ties (Carey
1990a; Carey 1990b; Leatherman 1996; Leatherman 1998b; Leatherman and others 1988), and the relative importance of undernutrition as a predictor of the delayed pattern of growth, outlined by other investigators (Carey 1990a; Leatherman 1994; Leonard and others 1990). I will discuss each of these in turn.
As in other regions subject to a high degree of seasonality, households in the rural Andes experience increased risk during the planting season, when food stores are at their lowest, and labor requirements at their highest (Branca and others 1993; Ferro Luzzi and Branca 1993;
Johnston 1993; Leatherman 1998b; Leatherman and others 1988; Leonard 1989b; Leonard 1991;
Leonard and Thomas 1988; Valverde and others 1982). Overall, agricultural production in the
Andes is characterized by low crop diversity, low yields, and unpredictability in yearly productivity (Leonard and others 1990). While these ecological constraints are critically important to the survival of all organisms, and require an adequate response, poverty represents the most important determinant of the degree to which they influence human biology
(Leatherman and others 1988; Leonard and others 1990).
Contrary to their predictions, these investigators found that reciprocal labor exchanges, a typical means of dealing with seasonality in the Andes, were relatively uncommon (Carey 1990a;
Leatherman 2005). Many cultural anthropologists have written about the importance of ayni , a system of labor exchange between kin and fictive kin, which helps ensure that households have access to enough labor during critical agricultural periods (Bastien 1985; Buechler and Buechler
1971; Weismantel 1988). As in other parts of the world, the commoditization of labor has led to
64 a decline in reciprocal relationships, at the same time that inadequate wage earnings necessitate continued reliance on this type of support (Collins 1986; Collins 1988). In Nuñoa, there is an increasing tendency to exchange labor exclusively on the basis of monetary payment, even between parents and children (Leatherman 1998a).
These investigators found illness, and effect of illness on household reproduction, to be unequally distributed. That is, the contribution of illness, mainly in the form of musculoskeletal pain, headaches, and respiratory infections, to reducing the labor power of households enhanced existing inequalities (Leatherman 1996). Women heading their own households, though they experienced a similar frequency of perceived illness episodes, lost more work days to illness than women in dual headed households, perhaps due to an increased burden of labor on individuals within smaller households (Leatherman 1996; Leatherman 1998a). In addition, those households whose members were chronically ill were less likely to have access to extra household labor because others were reluctant to enter into labor exchanges that might not be reciprocated
(Leatherman 2005; Leatherman 1994; Leatherman 1996). In this way, these authors demonstrated the dialectical relationship between illness and poverty: households with a smaller labor pool and subject to more chronic illness planted fewer crops, and found household reproduction increasingly challenging over time (Leatherman 2005; Leatherman 1996;
Leatherman 1998b).
Several investigators have confirmed that high altitude hypoxia does contribute to delayed growth and short adult stature among Andean peoples (Greksa and others 1985; Stinson
1982), however; Leonard and colleagues (Leonard 1987; Leonard 1989a; Leonard and others
1990) were the first to demonstrate the relatively greater importance of undernutrition in shaping growth and development in this setting. In contrast to previous dietary studies, Leonard (1987)
65 explicitly examined the importance of seasonality in consumption patterns. He found that, while there are no differences in dietary intake during the post harvest season, lower income families had significantly lower calorie and protein intakes during the pre harvest season due to their inability to make up for shortfalls through the purchase of market foods (Leonard 1987; Leonard
1989a; Leonard and others 1990). In addition, dietary quality had improved for individuals of high socioeconomic status since the 1960, but this was not the case among poorer households
(Leonard 1989a). These dietary patterns were reflected in growth and body composition, as individuals from wealthier households were significantly taller and heavier and had uniquely experienced secular trends in growth since the Nuñoa period (Leonard 1987; Leonard and others
1990).
These authors explored potential household strategies for dealing with seasonality in diet, labor requirements, and illness. Though it was less common than they expected, some households, particularly those located in a rural allyu , relied on instrumental social support to increase their labor pool (Carey 1990a; Leatherman 1998b). Other strategies included reliance on the labor of children, who have lower caloric requirements for the same work, facilitated through increased allocation of food to children during seasonal shortages, the emigration of older males during times of resource shortage to reduce household energy needs, and the participation of adults in sedentary activities to decrease individual energy demands (Leonard
1991; Leonard and Thomas 1989). This is confirmed by the finding that children from households with more dependents, and therefore more access to child labor, were significantly taller (Carey 1990a).
This project made several important contributions to understanding human biology in the
Andes and to the discipline more broadly. First, these authors demonstrated a sophisticated
66 understanding of the concept of adaptation, suggesting the importance of various levels of response, without implying that they were either entirely successful, or cost free. Second, they helped illustrate complex interactions between the biotic and social environments, and the dialectical relationship between illness/undernutrition, and poverty. Finally, these authors have created a foundation from which to expand political economic perspectives on human biology, particularly within the Andes. Leatherman and Carey, both advocates of political economic approaches, though they do not explicitly discuss how political economy might change how we think about culture; their work provides important insights into the influence of capitalist development on health in the Andes. Investigators from this project have convincingly made the case that much of what was once naturalized as an inevitable result of life at high altitude, is contingent upon the local effects of global processes. This conclusion is supported by much the research focused on high altitude physiology, which suggests that Andeans are well adapted to hypoxic conditions (Vitzthum and others 2000; Vitzthum and Wiley 2003).
The same large scale forces that contribute to undermining rural livelihoods in the Andes are related to rapid urbanization throughout the global south, and the consequences of this for human biology are still poorly understood. Anthropologists, in general, have done relatively little work in urban areas and are underrepresented in urban studies journals and departments
(Low 2002). Biological anthropologists, in particular, are increasingly interested in the relationship between urbanization and health, but only beginning to grapple with the complexity of and variability of urban environments. In the following section I will discuss the work of biological anthropologists on human biological variation within cities in the developing world.
Urbanization and health
67 According to the United Nations, 30 percent of people lived in cities in 1950; this year,
2008, will mark the point when half of the world’s population lives in urban areas, and by 2050 this proportion is projected to reach 70 percent (UN 2008). In the coming decades, nearly all population growth will take place in cities in the global south, and much of this will involve the rapid expansion of urban slums and shanty towns (UN 2008). In contrast to many parts of Asia and Africa which are still largely rural, cities in Latin America experienced enormous growth over the last century, and this region now has a lower proportion of people living in rural areas than Europe (UN 2008). In order to understand variation in health in the developing world, it will be increasingly important to understand its relationship with urbanization. At the same time, while urban migration is, at least in part, a consequence of neoliberal globalization (Farthing and others 2006), it can only be fully understood as an outgrowth of local dynamics. In this section, I will review some of the research on health among urban residents and migrants, and discuss how this work can be expanded to incorporate a more global focus.
In general terms, individuals living in urban areas enjoy improved health, along a number of dimensions, when compared to those living in rural settings (Sastry 1997). In Latin America, cross country comparisons indicate that mortality is, more often than not, higher in rural than urban areas, particularly under the age of two (Sastry 1997). In Brazil, urban mortality exceeded that of rural areas in the 1960s and 70s, but this pattern has become reversed in recent years
(Sastry 1997). There are also some aspects of health that tend to decline with urban migration.
Schell (1997) argues that exposure to infection and toxic substances can be higher in an urban environment because of the concentration of people and industry. In addition, chronic health problems associated with declining activity levels and the increased consumption of calories from highly processed foods, are more of a concern for populations in cities (Pelto and Pelto
68 1983). In this way, culture, in addition to potentially serving as a buffer, also enhances stress
(Schell 1997). Despite these general patterns, rural/urban comparisons mask an enormous amount of variability in health in both settings, and anthropologists have become increasingly concerned with investigating the heterogeneity of urban environments (McDade and Adair 2001;
Schell and others 1993a; Szwarcwald and others 2002).
A large number of factors have been proposed to account for these overall differences in health in these settings. Rural areas often lack services and resources that are more common in urban settings including health care and education facilities, access to wage labor, and sanitation.
In the rural Bolivian highlands, the number of health facilities is inadequate, and the rough terrain prevents some villages from having access to those that do exist (Perry and Gesler 2000).
By extension, individuals migrating from rural areas often have less education and fewer resources than those living in a city, which may confer disadvantages in terms of access to employment (Sastry 1997; Stephenson and others 2003).
Of particular concern is understanding the health effects of the rapid expansion of urban slums within the developing world (Adair and others 1993). Neonatal mortality in urban slums in India is higher than better off urban centers, and is similar to that seen in rural areas
(Fernandez and others 2003). Panter Brick and colleagues (2001; 1996) undertook a study of the health of Nepalese boys living in different contexts within the urban setting of Kathmandu, including middle class school children, individuals living in a squatter settlement and those living on the street, as well as a group living in a rural village. Not surprisingly, the school boys were found to be taller than any of the others, but the homeless boys were significantly taller than the squatter boys, and both were significantly taller than the boys from the rural village
(Panter Brick and others 1996). Plasma levels of ACT were substantially higher among the
69 village boys than among any of the urban children, followed by the urban homeless boys, the urban squatters and the middle class boys, and all inter group differences were significant
(Panter Brick and others 2001).
Another area of study has been health differences between migrants and residents or between long term and more recent migrants in urban settings. In India, under two mortality is higher for rural than urban residents, and higher for migrants than non migrants, a finding explained largely by differences in socioeconomic and demographic factors (Stephenson and others 2003). In a comparison of a number of developing countries, Brockerhoff (1995) found that urban migrants tend to be disproportionately female and that they experience greater child mortality than established city residents. Similarly, the height of children born in Guatemala
City differs based on the migration status of their parents (Bogin and MacVean 1981). Children with two city born parents were taller than children with one migrant and one city born parent.
In Kathmandu, mean HAZ and WAZ scores were significantly higher for children whose mother were long term migrants, than for children of more recent migrants (Moffat 1998).
A number of authors have also investigated the importance of smaller scale variation in neighborhood characteristics for explaining health within urban settings. Szwarcwald and colleagues (2002) found that residential concentration of poverty, as measured by the mean monthly income of household heads in each area, was predictive of increased infant mortality.
These authors suggest that this pattern can be explained by inadequate access to infrastructure, and lower quality housing. In Cebu, peri urban areas offer significantly less infrastructure, including electricity, roads, public transportation and health care facilities, than the urban core and squatter settlements (McDade and Adair 2001). Recent migrants and women tend to be overrepresented in more marginal urban areas (Brockerhoff 1995).
70 In addition to differences in infrastructure, cities offer greater opportunity for wage earning activities, whether in formal or informal sectors of the economy, which, some argue, lead to relative improvements in health, particularly for women and children (Bisgrove and Popkin
1996). In Cebu, informal work, in particular, was beneficial for women’s nutritional status and dietary quality, as well as that of their children (Bisgrove and Popkin 1996; Ukwuani and
Suchindran 2003). Short term and flexible work may be of particular benefit to mothers, as it often allows them to take their children to work and continue to breastfeed infants (Bisgrove and
Popkin 1996; Dufour and others 2003; Moffat 2002; Ukwuani and others 2001).
In relation to rural areas, urban settings tend to be associated with an increased quantity and variety of foods. Panter Brick and colleagues (1996) argue that the relatively better health of homeless boys in Kathmandu relates to a combination of reduced energy expenditure than seen in rural villages, and a more varied diet. Women living in poor barrios in Cali have adequate nutritional intake as measured by anthropometric indices and dietary intake, and employ a number of different strategies for dealing with shortfalls (Dufour and others 1997). These include relying on relatives or friends, store credit, charity or social programs, and altering their consumption by limiting protein quantity and meal sizes. In these, as in other settings, poor urbanites spend a high proportion of their income on food (Dufour and others 1997; Panter Brick and others 1996). In addition, as is true in the US, food insecurity in urban areas is increasingly associated with obesity as well as undernutrition (Brown and Konner 1987; Fitchen 1988).
A number of studies have focused on individual engagement with modern medicine. In
Karachi, Pakistan, child mortality is argued to be associated with mothers’ seeking treatment for their children, but also which type of treatments they choose (D'Souza 2003). This author found that using traditional forms of medicine, changing healers quickly, and failing to understand
71 medical instructions were all negatively associated with child survival. Increased child mortality among urban migrants in India was related to treatment seeking behavior in a national study
(Stephenson and others 2003). In contrast, Moffat (1998) found that differences in the height of children of migrants and non migrants in Kathmandu were not explained by treatment seeking behavior, and that women were not more likely to seek medical treatment if they had lived in the city longer. In fact, she found that access to medical care was one of the reasons women migrated to this urban center in the first place (Moffat 1998). In addition, there are many potential drawbacks for urban migrants in seeking medical treatment. Ayora Diaz (1998) argues that people often become dissatisfied with western medical practices because they focus narrowly on the individual systems within the body, and don’t consider the importance of social context. Crandon Malamud (1991), suggests that western medicine, as it does not address poverty, and the ultimate causes of poor health, legitimately fails to inspire confidence among local populations.
Associated with urbanization in the developing world is an increase in the proportion of households headed exclusively by women, particularly in South and Central America (De Vos and Richter 1988; Kennedy and Peters 1992). One of the reasons for the proliferation of research regarding female headed households (FHH) is that they are often economically disadvantaged compared to male or dual headed households (MHH, DHH) and are more likely to be found in low income neighborhoods (Ahsan 1997; Blumberg 1993; De Vos and Richter
1988; Johnson and Rogers 1993; Kennedy and Peters 1992; Ray 2000; Shell Duncan and Obiero
2000; Staten and others 1998). There are a number of potential reasons for the economic disadvantage of single mothers: 1) females are often at a disadvantage in acquiring adequate housing because of male bias in financing and inheritance (Deere and Leon 2003; Klak and Hey
72 1992); 2) women face discrimination in the labor market, and often occupy the lowest paying jobs (Ahsan 1997; Brown and Lapuyade 2001); 3) female household heads have to balance child care with economic activities, often with great difficulty (Ahsan 1997; Blumberg 1993; Muthwa
1994); 4) and FHH are likely to have fewer economically active individuals and therefore to be taking in less income relative to the number of dependents (Ray 2000).
Though categorizing FHHs is to some extent arbitrary, as is categorizing households in general, it involves the recognition that these households take diverse forms and allows for statistical comparisons (Kennedy and Peters 1992). Female household headship can be de facto, in which case the husband is absent more than half of the time, though he may still contribute some earnings to the household (Kennedy and Peters 1992). Alternatively, de facto FHH can be characterized by having an adult male present, but one who is unable to contribute economically to the household (Staten and others 1998). De jure FHH are those households in which a woman is the legal head of the household and controls most or all of the household assets and income, whether unmarried, widowed, or divorced/separated (Kennedy and Peters 1992).
A number of studies have found that, despite generally lower socioeconomic status, women and children living in FHHs in the developing world enjoy nutritional status that is equal or superior to those living in MHH or DHH (Blumberg 1993; Johnson and Rogers 1993;
Kennedy and Cogill 1987; Rogers 1996; Staten and others 1998). In FHH in the Dominican
Republic, children were better nourished despite the fact that the per person household energy intake was lower than among DHH (Johnson and Rogers 1993). Kenyan preschoolers living in de facto FHH had better nutritional status than those living in de jure FHH or MHH, despite having the lowest income of all the household types (Kennedy and Peters 1992). This study highlights the need to examine the variation in nutritional status within different types of FHH.
73 Among the Rendille, the nutritional status of children in FHH was significantly better when the income of the family was high, but the reverse was true when family income was low (Shell
Duncan and Obiero 2000). These researchers attribute this finding to the tendency for FHH to be among the poorest families in the community; pointing to the probability that women’s coping mechanisms are likely to be ineffective below a certain resource availability threshold.
The pattern of relatively good nutrition in FHH as compared to MHH has largely been attributed to nurturing behavior of women, including a tendency to divert a larger percentage of household income into food (Shell Duncan and Obiero 2000; Staten and others 1998), to divert a greater proportion of the food coming into the household to children (Johnson and Rogers 1993;
Kennedy and Peters 1992), to invest in higher quality foods such as from animal sources (Rogers
1996). An additional possibility is that women may rely on support from kin to make up for shortfalls in resources (Staten and others 1998).
Fewer studies have investigated differences in rates of morbidity between FHH, MHH and DHH. In addition, these studies have relied on self or caregiver reported morbidity rather than objective measures. In the Dominican Republic, children from FHH, particularly from the lowest income quartile, were more likely to be sick at the time of the interview than children from MHH (Johnson and Rogers 1993). Despite a relatively large difference in rates of morbidity, child growth did not appear to be negatively affected. The authors attribute this to improved care of sick children in FHH, but suggested that female household heads were more likely to report their children as sick than members of male headed households. In contrast,
Kenyan preschoolers from both de jure and de facto FHH were less likely to have diarrhea and other diseases than children from MHH (Kennedy and Peters 1992). These researchers attributed the lower rates of stunting seen among FHH partly to lower rates of infectious disease.
74 There are several reasons why morbidity might be expected to be higher among children from FHH including that exposure to pathogens tends to be more common in lower income neighborhoods (Solomons and others 1993). However, improved nutritional status may protect children from infection to some extent (Kennedy and Peters 1992). It is important to adequately consider nutritional status and disease load together, because of the synergistic relationship between infection and under nutrition (Pelletier 1994).
While most of the focus on FHH and health is related to their economic status, this is not always the case. In a recent article based on research in Jamaica, Bronte Tinkew and DeJong
(2004) hypothesized that two parent households would be associated with the best health for children, with cohabiting, extended family, and single parent households contributing to reduced growth. With regard to cohabitation in particular, these authors argue that “while the household may benefit from the additional income, cohabitation is not a substitute for marriage” (Bronte
Tinkew and Dejong 2004: 501). Their expectations were largely confirmed, with children from two parent households demonstrating the highest average HAZ. These authors consider the relationship causal, and do not consider the possibility that poor health and household structure are both related to class or income or that household structure might be a response to these or other factors.
These studies have provided some important insights regarding urban health in the global south. While it is important to recognize that individuals living in urban environments tend to be better off than those living in rural areas, this is relative, and there is an enormous amount of individual and neighborhood level variation. For example, Panter Brick and colleagues (1996: page) argue that “the adoption of street life may, for some at least, represent both a rational and successful response to prior circumstances.” While it is likely that this response is rational for
75 some boys in Nepal, success must be seen in relative terms. Though the homeless boys were relatively well off in relation to boys living in rural villages, they were still relatively worse off in comparison to middle class children and US controls.
A number of these studies have also highlighted some of the problems associated with the biomedical paradigm within which they were framed. In several pieces, the explanatory framework involves a ‘culture of poverty’ style argument, in which individuals are thought to rely on cultural norms that are inconsistent with successful livelihoods (Chin 2001). While the focus of these articles is on improving public health, the explanations of health differences are based ultimately on differences in individual qualities, presumably based on individual choice, including educational attainment, occupation and area of residence, health seeking behavior, and marriage status. In the public health literature, in general, rationality is often determined based on the degree to which the behavior of the subjects conforms to western cultural norms. In their discussion of residential poverty clustering, Szwarcwald and colleagues (2002: 2089) argue that teen pregnancy and single parent families are some of the most important public health concerns in Brazil and that “the goal is the individual behavioral change which in turn depends on the transformation of deeply ingrained social norms.” Fernandez and colleagues (2003: page) suggest that child mortality is a result of “illiteracy, ignorance and poverty resulting in women not paying attention to pregnancy and health.” A number of authors provide similar arguments suggesting that ‘traditional’ beliefs are a hindrance to the assimilation of migrants to the urban environment, something with negative consequences for health (Brockerhoff 1995; Stephenson and others 2003; Szwarcwald and others 2002).
This focus on cultural and individual explanations places the responsibility for poor health and poverty on the individual actor (Wood and Salway 2000). In general this represents a
76 western bias in seeing the individual as subject to few constraints in economic decision making, with rationality defined as behavior that seeks, always, to maximize economic efficiency (Carrier
1997). As I discussed in chapter one, this bias is present not only in research, but in the interventions directed by NGO’s and international organizations, and so can have profound consequences for naturalizing health inequalities. Armelagos and colleagues (1992: 42), have argued that “options for response to insult are not infinite. They are constrained by wealth, class position, social relations, and ideology.” Anthropologists, with our reliance on in depth ethnographic methods and emphasis on cultural relativity, are well positioned to understand the local rationality of economic and health related decision making, but I argue that political economy is also necessary for this purpose. Urbanization in the global south is dependent on unique combinations of large scale and local forces. For example, the implications of urbanization for health are likely to be very different in the developing world, than for Europe and North America, because the drive toward urbanization is based on different historical and political currents (Schell and others 1993a).
El Alto
77
Figure 2.1. Map of El Alto and La Paz, including major route in and out of the cities.
In the global south, neoliberalization is associated with the depopulation of rural areas, as the state retreats from regional involvement and private industry outcompetes small scale farmers, and leads to the rapid expansion of urban slums and squatter settlements (di Leonardo
2008). At the same time state investment in these marginal urban communities remains low
(Goldstein 2004; Lazar 2008), and/or focused on their consolidation and partition from wealthier enclaves (Caldeira 2002; Gledhill 2005). Individuals living in peripheral urban communities must rely on informal or formal associations, such as neighborhood or labor organizations to demand rights of citizenship from the state, which may have profound implications for their ability to mount a resistance to neoliberal policy (Goldstein 2004; Lazar 2008). In this section, I
78 will discuss the growth and economic realities of El Alto, and finish with a discussion of the unique social organization of this city.
The majority of alteños are indigenous migrants from rural areas on the altiplano, a vast plane, roughly 4000 m in altitude. Related to colonial rule in Bolivia is a history of intense racial and gender discrimination, which have direct implications for the production of economic inequality (Gill 1994; Weismantel 2001). This is well illustrated in La Paz El Alto, where indigenous people tend to live in the most impoverished areas and have few options with regard to employment. A number of authors have remarked that altitude in La Paz (population
1,576,100 metro. area, 830,500 city proper) the administrative capital of Bolivia, is a relatively accurate indicator of wealth, ethnicity, and access to resources and services (Arbona and Kohl
2004; Gill 1994). The lowest altitude section of La Paz houses Zona Sur , a wealthy suburb inhabited by a large number of wealthy foreigners and individuals who label themselves as white or criollo (Gill 1994). Fiver hundred meters higher is the downtown core of La Paz, set within a steep basin. The poorest in La Paz reside in the most precarious spaces on the walls of the basin and face the periodic threat of mudslides. At 3800 m above sea level, on the altiplano and overlooking the city of La Paz is El Alto (population 728,500), a city primarily occupied by indigenous Aymara people from the surrounding countryside (Gill 2000). El Alto has grown to rival La Paz in size, and was made a separate municipality in 1988, but this masks the continued, unequal relationship between the two cities (Gill 1994; Pacheco 1997).
El Alto, depicted in relation to La Paz in figure three, does not have a city center like other municipalities. The center of commerce, transportation between the cities, labor and neighborhood organizations, and municipal power resides in the Ceja (eyebrow), overlooking the city of La Paz (Lazar 2008) (See figure 2.1). Routes between La Paz and other municipalities
79 such as Oruro, Cochabamba, and Copacabana run through El Alto, giving protestors a strategic advantage in terms of their ability to block both commercial and tourist traffic, something which may have influenced the relative success of major protest actions centered there (Dangl 2007).
El Alto experienced its first major period of growth in the 1950s, starting with a population of roughly 11 000, with agrarian reform and the dismantling of the hacienda system
(Arbona and Kohl 2004). Though the population increased at a relatively constant rate, major waves of migration were stimulated by the mine closures and drought in the mid 1980s. Semi proletarianization and labor migration mean that people no longer have time to invest in the social ties critical for recruiting extra household labor or less important agricultural task, a problem in the unpredictable and marginal agricultural environment of the altiplano. Gill (2000) argues that people come to El Alto out of desperation, rather than because they are drawn by the possibility of employment. El Alto houses some limited manufacturing facilities, but by far the majority of individuals in El Alto are employed in the informal sector (Arbona and Kohl 2004).
Many work in La Paz as day laborers, shoe shine boys, prostitutes and market vendors or as domestic servants for the wealthy. Indeed, La Paz depends on the cheap source of labor provided by indigenous people from El Alto (Arbona and Kohl 2004; Gill 1994).
Although potentially subject to less state control, the informal sector is fundamentally related to the formal sector in capitalist economies, for example, by the expansion of subcontracting as a means of reducing labor costs and increasing flexibility in the formal system
(Leonard 2000; Sassen 1998; Sassen 2001). Non capitalist forms of labor benefit capitalism because they lower the cost of labor and increase surplus value (Long and Richardson 1978).
The tendency of capitalist development is to create large numbers of workers who are unprotected by formal labor standards, and easily exploitable, a process which tends to
80 differentially affect individuals by gender, age and race (Leonard 2000). Indeed, the very young, the very old, and women are usually over represented in this sector of the economy (Long and
Richardson 1978), though men are increasingly forced to turn to informal employment as economic conditions worsen (Seligmann 2004). The informal economy is no less important than in western countries, and likely fulfills many of the same functions including providing low cost goods and services to a growing proportion of low income workers (Sassen 1998). In addition, informal does not necessarily mean unregulated. In El Alto, informal markets are managed by labor federations such as The Federation of Street Traders, and the Regional Workers Center, through which workers can band together to form a regulatory structure of benefit to all (Lazar
2008). In this context, personal and family connections are important for determining things like entry to the market and access to prime locations (Leonard 2000; Peña 1999).
While La Paz has experienced relatively measured growth, the rapid explosion of migration into El Alto has meant that municipal infrastructure is unable to keep pace. In addition, La Paz takes in 5 times more revenue than El Alto, partly because businesses operating in El Alto are more likely to be headquartered in, and pay taxes to La Paz, and partly because the informal nature of most employment restricts the tax revenues that can be generated from individuals (Arbona and Kohl 2004). This means that a smaller number of households in El Alto have access to amenities such as plumbing and sanitation (Arbona and Kohl 2004). Despite the additional income that came with the Law of Popular Participation, the operating budget of El
Alto is unable to meet the demands of its citizens, and corruption of local government remains a major problem (Arbona and Kohl 2004). It is perhaps for these reasons that public participation in the economic affairs of the city remains relatively low (Merkle 2003).
81 Access to resources also varies a great deal within this city. There are few hospitals operating in El Alto, and the distribution of health clinics is uneven (Gill 2000). NGOs pick up some of the slack, but are generally located in the center of the city. Those having arrived more recently, and living on the outskirts of the city are less likely to be able to access these services.
Similar issues exist with regard to schools. With El Plan de Todos, the city became responsible for both building and staffing new schools, which has led to serious problems of overcrowding and lack of resources in the classroom. A small number of private schools have begun to fill the void left by the public system, but most Alteños cannot afford to send their children to these institutions (Gill 2000). Housing is also a major issue for immigrants to El Alto. Land lord/tenant relationships are often exploitive, and most Alteños must struggle for years to own their own home (Gill 2000). Obtaining lots is difficult and loteadores often raise their profits by selling the same land to more than one family, selling fictive lots on public land, or failing to provide promised services (Gill 2000). Once families have come up with the means to purchase land, they often have to labor long hours to buy materials, and spend any remaining time on construction, a process which often takes years (Gill 2000).
There are two major forms of organization in El Alto, mediating between local residents and the state: the first are neighborhood organizations, or juntas vecinales , organized into the larger federation known as FEJUVE; the second are labor organizations, discussed above, which work to resolve conflict between, for example, individual vendors, and vendors and the state
(Lazar 2008). In both cases, these organizations are somewhat obligatory, for example, families are fined for not coming to demonstrations or meetings, and some level of involvement is generally necessary for commercial or neighborhood success (Lazar 2008). Much of state involvement in El Alto, for example infrastructural investment, is mediated through these
82 organizations (Lazar 2008). The FEJUVE was instrumental in organizing against water privatization in El Alto, which substantially raised the price of water, while leaving large portions of the city out of plans for infrastructural development (Laurie and Crespo 2007). Lazar argues that the Federation (of street traders), as a collective bargaining association, and a mediator between local street vendors, wholesalers, and rural producers, helped to enhance solidarity among different social movements, both rural and urban, during some of El Alto’s larger scale protests, and helps to explain their success (Lazar 2008).
Gill (2000) argues that migrating to El Alto is associated with a lot of stress. There is much competition for limited resources in this city, and gender relationships within the household often become strained as women enter the labor force and families confront economic hardship (Gill 2000). Survival in El Alto requires a diversified economic strategy and, potentially, reliance upon networks of support during particularly difficult times (Gill 2000;
McFarren 1992). Their economic vulnerability may make this particularly important for women who head their own households. Further, involvement in supportive networks may mitigate, to some extent, the negative effects of economic inequality on health.
Conclusion
In this chapter I have outlined the development of political economic approaches to human biology in the rural Andes, some of the specific forces which undermine rural livelihoods and drive urban migration, particularly in Bolivia, and some of the health consequences associated with this trend. Though biological anthropologists have increasingly focused on the complexity of urban environments, particularly in terms of the localization of large scale forces, there are still a large number of academic studies which take for granted neoliberal development as the key to reducing health inequalities. Anthropologists are in a unique position to investigate
83 the political economic forces which constrain household production, the coping mechanisms used by households and individuals to deal with these constraints, and the physical embodiment of economic and political inequality in the form of health outcomes. In the next chapter, I will examine the physiological pathways linking social support and health, along with some of the discussion of the need to clarify the importance of large scale phenomena in influencing this relationship.
84 Chapter 3: Social relationships and health
Introduction
In this chapter I will examine evidence for social support, and social relationships in general, as predictors of physiology. Much of the evidence for this relationship comes from investigators working in social epidemiology and psychology, and as such, diverges somewhat from the concerns of anthropology (Dressler and others 1997; Dressler and others 1986;
Jacobson 1987). However; this work is important for establishing the mechanisms by which social relations are embodied. In the first section I will review evidence for the influence of social support on endocrine, cardiovascular and immune function, and nutritional status, and hypotheses designed to explain this influence. In particular, I will focus on the common mechanisms linking the social environment to each of these systems, and on their interactions with each other. In the second section, I will focus on current approaches to conceptualizing and measuring social support, and in particular, anthropological approaches to understanding the cultural context of social support.
Social support
A large number of authors trace the beginning of a rapid proliferation of research on social support and health to the work to two epidemiologists (e.g. House and others 1988):
Writing independently in the same year, Cassel (1976) and Cobb (1976: 300) both argued that social support, or “information leading the subject to believe that he is cared for and loved, esteemed, and a member of a network of mutual obligations,” acts to buffer the negative effects of life stress on human health, whether from universal age related transitions, or that associated with illness, bereavement and other events. Much of the data available at the time were far more suggestive than conclusive, but the work of these authors sparked a wave of investigation; in the
85 early stages, particularly, on the relationship between social integration, a count of individual’s social contacts, and mortality (House and others 1988). In a widely cited article, compiling information from prospective studies, House and colleagues (1988) demonstrated a significant and consistent link between fewer social contacts and increased risk of all cause mortality. In particular, these authors argued that low social integration constitutes a predictor of mortality similar in effect size to smoking, controlling for other risk factors. The mechanisms behind this association, however; were still largely unexplored.
There are a number of aspects of social relationships that are of interest to scholars in diverse fields. There is a clear relationship between the number of social contacts that a person has, as well as the number of roles that person occupies in relation to others, and certain biological outcomes (Cohen and others 1997; House and others 1988). Another area of interest, particularly to anthropologists, has been understanding the characteristic patterns of social networks in different societies, and the influence of these patterns on specific individuals, referred to as ego centric network analysis (Berkman and others 2000). Recent studies have focused increasingly on the qualitative aspects of social relationships and the benefits derived from them, specifically, social support (Berkman and others 2000). Social support is often classified based on the type of aid it confers, for example Franks and colleagues (2004) suggest that social support can be classified into four categories: 1) emotional social support, or a relationship that reinforces feelings that one is cared for and belongs; 2) instrumental social support, or aid in labor or money; 3) appraisal social support, or access to counsel regarding important decisions, and; 4) informational support, or relationships which serve as an important source of data, for example, on health.
86 There are two major, and partially competing, hypotheses on the relationship between social support and health. The first, the direct effects model, argues that social support has a positive influence on health, and enhances numerous aspects of physiology. The buffering model, in contrast, argues that social support acts primarily to buffer the negative effects of stress, whether by changing people’s appraisal of stressful events, or some other mechanism
(Cassel 1976; Cobb 1976; Cohen 1988; Cohen 2004; Cohen and Wills 1985). In this second case, one would expect to observe a relationship between social support and health only under stressful circumstances (Cohen and Wills 1985). In order to directly test the buffering hypothesis, Uchino and colleagues (1996) suggest that stress must be tested as a significant predictor of the dependent variable, and this relationship must be substantially or entirely attenuated with the addition of social support to the model. Cohen has argued widely that social integration acts on health directly, while social support acts primarily through moderation of stress (Cohen 1988; Cohen 2004; Cohen and Wills 1985), though others have suggested that these pathways are not mutually exclusive (Miyazaki and others 2005). An alternative that has recently received a great deal of attention is that social isolation and negative social interactions are, themselves, important sources of stress (Hawkley and others 2006; House 2001; Knox and
Uvnas Moberg 1998; Nausheen and others 2007; Newsom and others 2003).
There are several potential confounders of the relationship between social support and health. One is the possibility that much of the relationship between these variables can be explained by health behaviors. That is, individuals with more social support tend to have a better diet, engage in more physical activity, and other behaviors that enhance well being (Uchino and others 1996). Data from a large number of studies support the conclusion that, while these behaviors may sometimes be a factor, they generally do not explain much of the relationship
87 between social support and health (Uchino and others 1996). The second, of particular concern among psychologists, is that aspects of personality, for example hostility, may affect both access to social support and health (Uchino and others 1996). Again, personality characteristics do not appear to fully account for the effects of social support on health, but may sometimes interact with social support in predicting biological outcomes (Angerer and others 2000; Knox and others
2000; Ong and Allaire 2005; Ratnasingam and Bishop 2007; Westmaas and Jamner 2006; Wirtz and others 2006). Finally, a number of researchers have expressed concerns about the direction of causality. For example, there is evidence that mental illness is more likely with reduced social support, but that social support also tends to decrease when individuals begin to show signs of mental illness (Dickinson and others 2002). This suggests that causality may be bi directional.
Stress, adaptation and sociality
During the late nineteenth century Claude Bernard, often described as the father of investigative medicine, researched, among other things, the role of the pancreas in digestion, the liver in the synthesis of sugar, and the effects of curare on the nervous system (Tan and Holland
2005). Based on his physiological experiments, he proposed the concept of le milieu intérieur , arguing that the internal environment, bathed in interstitial and circulating fluids, is internally regulated to resist fluctuations in the external environment (Goldstein and Kopin 2007;
Normandin 2007; Tan and Holland 2005). Following up on this work during the early twentieth century, another experimentalist physician Walter Cannon, introduced the concept of homeostasis to describe the idea that the body maintains tight regulation of internal conditions, for example pH and blood pressure, using negative feedback mechanisms (Goldstein and Kopin
2007; Sterling 2004). In addition, he uncovered the importance of sympathoadrenal activation as
88 a response to threats to homeostasis, labeling this phenomenon the “fight or flight” response
(Goldstein and Kopin 2007; Sterling 2004).
The term stress is generally used to describe the body’s response to emotional or physiologic challenge (McEwen 2007); while stressor refers to the environmental conditions that cause this reaction, specifically, any stimulus that endangers life, and must be met with an adaptive response (Selye 1950; Selye 1976). In the early fifties, Selye outlined the general adaptation syndrome (GAS), a systemic, non specific reaction, which he separated from any localized effects of environmental stressors (Selye 1950; Selye 1976; Selye and Fortier 1950).
Signals carried from the affected tissue, via the nervous or cardiovascular system, to control centers in the body, initiate the three stages of GAS: the Alarm Reaction causes excitement or depression of the nervous system depending on the intensity and duration of the stimuli, along with a predictable cascade of neuroendocrine changes in the hypothalamic pituitary adrenal axis; the Stage of Resistance, during which physiologic function is partially restored, but the ability of the nervous system to adapt to additional challenges is constrained, and; the Stage of Exhaustion, in which additional challenges result in death. He defined diseases of adaptation, including allergies, rheumatoid arthritis, diabetes and hypertension, as conditions brought about by excessive or inappropriate physiological response to perceived stressors (Selye 1950; Selye
1976; Selye and Fortier 1950).
The idea that these diverse conditions stem from the same underlying causes, at the time somewhat controversial (Selye 1950), is now widely accepted, conceptualized within the framework of stress or, increasingly, allostasis. The term allostasis, introduced by Sterling and
Eyer (1988), in contrast to homeostasis, refers to “stability through change.” These authors recognized that constancy, per se, is less important than survival and reproduction, and that many
89 systems fluctuate considerably based on both internal and external conditions. Regulatory mechanisms must allocate limited resources to bodily functions during times of rest and action, and these resources are, to some degree, loaned from one system to another as conditions change
(Sterling 2004). In order to maximize efficiency and flexibility in physiological response, these regulatory mechanisms must anticipate the most likely demands, based on past experience, and change internal conditions to match them (Sterling 2004; Sterling and Eyer 1988). For example, blood pressure, rather than representing a constant value, fluctuates considerably depending on levels of activity and arousal, both above and below average values. These average values, in turn, depend on the brain’s prediction of the most frequent demands, so the “set point” will be different for each individual, and may change considerably during their lifetime (Sterling 2004).
For those individuals subject to a higher frequency of stressful situations, higher resting blood pressure represents an efficient regulatory response (Sterling 2004).
Others have taken a somewhat different view of allostasis, arguing, like Selye, that conditions such as hypertension represent a failure of adaptation. McEwen (2000; 2004a; 2005a;
2005b; 2007) proposes that we use allostasis as a more precise means of describing the sometimes vague stress concept. He uses the term to describe the active process of maintaining homeostasis; specifically, those systems which are homeostatic, or subject to tight regulation, are maintained by short term alterations in allostatic systems (McEwen 2000). For example, it is the flexibility of the autonomic nervous system in regulating cardiovascular function, which contributes to the maintenance of precise values for temperature and blood oxygen concentrations. Allostatic load refers to the cumulative physiological burden, or wear and tear of allostasis, whether because stressful circumstances require frequent response or allostatic systems react inappropriately or continue to react once the stressor has ceased (McEwen 2000;
90 McEwen 2004a; McEwen 2004b; McEwen 2005b; McEwen 2007; Stewart 2006). Chronic activation of the sympathoadrenal and hypothalamo pituitary adrenal axes, leads to a cascade of complex, and interrelated pathological changes in multiple systems, increasing allostatic load through time (Stewart 2006).
Animal studies consistently demonstrate that social isolation, particularly in early life, leads to a range of negative health consequences, suggesting that this is an important source of stress across a broad range of species (Lewis and others 2000; Sanchez and others 1995;
Tuchscherer and others 2006; Wu and others 2001). Among species in which dominance hierarchies structure access to resources, social stress can be associated either with dominance or subordinance depending on factors such as social stability and relative reliance on violence or affilliative behaviors to maintain or contest status (Kaplan and others 1991; Manuck and others
1995; Sapolsky 2005; Shively 1998; Uno and others 1989; Williams and others 1994; Williams and others 1991). Among a number of species of non human primates, dominant individuals are more likely to be groomed or involved in coalitions, as submissive animals tend to use these as coping strategies to avoid or resolve conflict (Sapolsky 2005). While sociality may be more complex among humans, it is not unique to our species. Sterling (2004) argues that there is generally some mix of competition and cooperation among conspecifics, and that goal oriented behavior depends on emotions to focus both thoughts and physiological regulation to match the circumstances. For humans, reciprocal emotional understanding, based both on symbolic communication and subtle physical cues, is critical to both collaborative and competitive interactions, and so it is unsurprising that its absence increases vigilance and anxiety (Sterling
2004).
91 Most of the research linking social support and health involves aspects of the cardiovascular system, but there is a growing focus on the immune and endocrine systems. I will begin each of the following sections with an overview of the function of, and interaction between physiological systems. My aim here is not to provide an exhaustive review of all of their components, but to focus on those that have been investigated with relation to social support.
Following this, I will review evidence for the link between social support and various aspects of physiology.
92 Hypothalamus (forebrain): produces and secretes releasing and inhibiting hormones Central nervous system Oxytocin Brain Corticotropin releasing hormone: Spinal cord stimulates the release of ACTH from the anterior pituitary
Pituitary gland, anterior lobe: produces hormones to stimulate other endocrine glands Peripheral nervous Corticotropin (ACTH): stimulates system adrenal cortex to release Autonomic division: glucocorticoids (cortisol) Parasympathetic Pituitary gland, posterior lobe: stores and Sympathetic: secretes hypothalamic hormones norepinephrine Oxytocin: causes sedation, lowers blood pressure/cortisol
Adrenal gland, medulla: produces and secretes Epinephrine: raises blood glucose, Cardiovascular function fatty acids, increases heart rate/force of contraction Norepinephrine: constricts or dilates blood vessels Immune function: Adrenal gland, cortex: produces and secretes Th1/Th2 balance Cortisol: raises blood glucose, reduces inflammation Nutritional status: Waist to hip ratio
Figure 3.1. Physiological pathways of the hypothalamic pituitary adrenal, and sympathetic adrenomedullary axes, linking perceived social support to endocrine, cardiovascular and immune function (the colors highlight different pathways within the same organ, and their downstream products).
Social support and neuroendocrine function
Figure 3.1 presents a simplified diagram of the pathways associating the nervous, endocrine, cardiovascular and immune systems. The central nervous system (CNS) is made up
93 of the brain and spinal cord, while the peripheral nervous system (PNS) carries signals to and from the CNS, in communication with the rest of the body. The brain is the most important organ linking stress and adaptation, as the regulatory center evaluating which stimuli are stressful, and coordinating behavioral and physiological responses (McEwen 2007). The PNS is divided into two categories, the somatic division which controls skeletal muscle, and the autonomic division which controls automatic bodily functions, such as those associated with internal organs. The autonomic nervous system can be further divided into parasympathetic and sympathetic divisions, which work antagonistically. Parasympathetic nerves play a greater role when the organism is at rest, and diverts the body’s energy toward housekeeping tasks such as digestion. Sympathetic nerves take over during situations of excitement or danger, and prepare the organism to take action (Johnson 2001).
Sympathetic nerves carry signals from the hypothalamus, and other regions of the brain, to secrete the catecholamines, epinephrine and norepinephrine. These molecules act as neurotransmitters in the nervous system and hormones when released into the blood stream. The catecholamines are manufactured and stored by sympathetic neurons, ending in the adrenal medulla. Sympathetic activation causes the release of acetylcholine from preganglionic sympathetic nerves, which stimulates the release of catecholamines. These hormones regulate the cardiovascular system, and carbohydrate metabolism during times of stress or excitement, producing the “fight or flight” response (Johnson 2001). This cascade of physiological responses is sometimes referred to as the sympathetic adrenomedullary axis.
The hypothalamus, located in the forebrain, monitors the state of internal organs and physiological processes, and interacts with the pituitary gland to form one of they body’s primary neuroendocrine control systems. The hypothalamus manufactures hormones which are
94 stored and released from the posterior lobe of the pituitary, one of which is oxytocin. Primarily involved in stimulating target cells in the mammary glands and uterus during reproduction, oxytocin can also play a role in sedation, lowering stress hormones, pulse and blood pressure
(Knox and Uvnas Moberg 1998; Uvnas Moberg 1998). The anterior pituitary produces a number of hormones which play critical roles in maintaining homeostasis in different physiological systems. Corticotropin (ACTH), stimulated by ACTH releasing hormone from the hypothalamus, acts on the adrenal cortex to stimulate the production of glucocorticoids.
Cortisol, the most important of the glucocorticoids, promotes the breakdown of protein from muscle and conversion to glucose in the liver, and promotes glucose sparing to give priority to the brain over skeletal muscle. Cortisol rises in response to physical or emotional stressors, and low blood sugar. This cascade is referred to as the hypothalamo pituitary adrenal (HPA) axis.
Neuroendocrine processes are generally regulated by negative feedback mechanisms.
For example, elevated plasma cortisol acts directly on the hypothalamus to decrease the secretion of ACTH RH, and the anterior pituitary to decrease ACTH, while stressful circumstances or a reduction in plasma glucose have the opposite effect. These neuroendocrine cascades have downstream (and reciprocal) effects on the cardiovascular and immune systems, and play an important role in growth and development, and body fat distribution. I will begin by discussing evidence for the relationship between social support and endocrine function, and then move downstream to discuss these other systems.
Animal studies indicate that social stress is associated with increased glucocorticoid secretion, and that among other downstream effects, are consequences for neural function. For example subordinate female cynomolgus monkeys ( Genus, species ) receive more aggression and less grooming than dominant females, and demonstrate higher levels of cortisol (Shively 1998).
95 The hippocampus, involved in spatial, episodic and contextual memory, and regulation of the
HPA axis, as well as the hypothalamus, have receptors for glucocorticoids (McEwen 2007;
Sapolsky and others 1986). A post hoc, post mortem examination of male vervet monkeys with evidence of bite marks (socially subordinate) and gastric ulcers, found hyperplastic adrenal cortices, along with selective damage to the hippocampus (Uno and others 1989). In contrast, socially isolated rats had reduced corticosterone levels, but displayed reduced spontaneous electrical activity in the hypothalamic paraventricular nucleus (Sanchez and others 1995). This suggests that stress related elevations in the HPA axis can have damaging effects on certain aspects of neural function, and can speed age related degeneration (McEwen 2007).
Among humans, there are relatively few studies linking social support to endocrine function. Uchino and colleagues (1996), in a qualitative meta analysis of all studies completed before 1996, found that there is consistent evidence for a negative relationship between social support and catecholamines, but little evidence to support an effect on cortisol. However; these authors also suggest that measurement issues, specifically, diurnal variation in hormone secretions and a lack of standardization in collection methods, may have contributed to these negative findings. More recently, some studies found that more positive daily social interactions, or social support, predicted a reduced cortisol response to laboratory stress (Eisenberger and others 2007; Heinrichs and others 2003; Kirschbaum and others 1995), while another demonstrates no such association (Thorsteinsson and others 1998). Rosal and colleagues (2004) found a negative association between social support and cortisol levels both at baseline and after one year. In contrast, Evans and Steptoe (2001) found a positive relationship between work related social support and cortisol on the weekend.
96 A number of recent studies have examined the potential role of oxytocin as a buffer of both sympathetic and HPA activation, whether elevated through touch or positive social interaction (Knox and Uvnas Moberg 1998; Uvnas Moberg 1998). Grewen and colleagues
(2005) found that women with higher reported spousal support had higher baseline oxytocin levels, and that this hormone mediated the relationship between partner contact and a reduction of blood pressure and norepinephrine in a laboratory setting. There was no evidence of an effect among men, which the authors suggest, may indicate some sex differences in hormonal reaction to social contact. In another study of laboratory stress, men were administered nasal oxytocin or a placebo, and randomly assigned to receive social support from a close friend, or to remain alone (Heinrichs and others 2003). There was a marginally significant effect of oxytocin in buffering the cortisol response to stress, but those assigned to both the social support and oxytocin condition had significantly lower cortisol than those with social support or oxytocin alone (Heinrichs and others 2003).
In a unique study, Eisenberger and colleagues (2007) tested links between activity in different areas of the brain and cortisol secretion in response to laboratory social exclusion.
Individuals reporting better daily social interactions over the previous ten days had reduced cortisol elevation in response to social exclusion. These authors found two brain regions to mediate the relationship between social exclusion and cortisol increase while a region near the hypothalamus was related to exclusion or cortisol but not both at the same time.
Social support and cardiovascular function
The cardiovascular system serves a critical role in transporting oxygen and nutrients to cells carrying metabolic waste products, helping to regulate pH and temperature, and circulating hormones and components of the immune system, and the heart serves as a pump for
97 transporting blood to and from every cell in the body. Blood flows through two cardiovascular circuits: the pulmonary circuit carries blood from tissue through the lungs to become re oxygenated; the systemic circuit carries blood to and from the tissues, supplying oxygen to each cell. A heartbeat involves one sequence of almost simultaneous contraction (systole) and relaxation (diastole) within the heart valves. Blood pressure is the pressure of the blood against the vessel walls during heart contractions, and is determined by a combination of the rate and force of the heartbeat and blood vessel diameter, both of which are responsive to sympathetic activation. A large number of studies have focused on the relationship between social support and cardiovascular function, as cardiovascular disease is a leading cause of death in western countries, and increased heart rate and blood pressure are independently predictive of later disease (Uchino and others 1996).
Many early studies focused on the influence of perceived social support on resting blood pressure and heart rate. Uchino and colleagues (1996: 490) report that, of 28 correlational studies of social support and cardiovascular function, 23 reported a significant negative relationship, four reported no relationship, and one a positive association. A meta analysis of 21 studies with available data on blood pressure, revealed a significant combined test, with an average effect size of 0.08 (Uchino and others 1996: 490). In particular, a number of anthropological studies provide support for a negative relationship between social support, or cultural consonance in social support, and blood pressure (Dressler and others 1997; Dressler and
Bindon 2000; Dressler and others 1986; Janes and Pawson 1986).
Some of these correlational studies provide support for the buffering hypothesis. For example, among male white collar workers, job strain was predictive of higher blood pressure only among those with low social support (Guimont and others 2006). Similarly, instrumental
98 social support was predictive of reduced blood pressure only for poor African Americans in the rural South (Strogatz and James 1986). Other studies demonstrated age related elevations in blood pressure only among those with fewer social relationships. Hawkley and colleagues
(2006) found “normal” age related increases in blood pressure only among those reporting high loneliness, though they suggest that loneliness and social support have independent effects on physiology. That is, loneliness may be an independent stressor, and not simply a proxy for low social support. Uchino and colleagues (1995) found age related elevations in blood pressure only among women with low appraisal support.
Primary evidence for the buffering hypothesis of social support comes from studies of cardiovascular reactivity to laboratory stressors, comparing individuals with and without social support. These studies involve well validated means of inducing a stress reaction, often preparing and giving a speech, or performing mathematical calculations in front of a small audience or video camera. The source of social support varies in these studies: some protocols ask participants to bring a friend or spouse with them, some provide a source of support from among the study personnel, and some use a measure of pre existing support. On balance these studies, including two meta analyses (Thorsteinsson and James 1999; Uchino and others 1996), demonstrate that heart rate, and systolic and diastolic blood pressure tend to be less elevated, in response to laboratory induced stress, in the social support condition (Christian and Stoney 2006;
Clark 2006; Kamarck and others 1990; Lepore 1995; Nausheen and others 2007; Ong and Allaire
2005; Ratnasingam and Bishop 2007; Treiber and others 2003). Uchino and colleagues (2001) found that individuals with few supportive ties demonstrated age related increases in heart rate reactivity to laboratory stressor, while the association was not present for individuals with high support. Many of these authors suggest that increased reactivity to stressful situations, over time,
99 leads to higher resting blood pressure, and other measures of chronic illness, but thus far, there is little evidence to support or contradict this hypothesis (Thorsteinsson and James 1999).
There are, however; cases in which the effect was significant, but opposite of the predicted direction. A number of studies have found increased cardiovascular reactivity in the support condition. Laboratory induced stress is produced by the prospect of evaluation, and a number of investigators have proposed that sources of support that are seen as potentially judgmental are likely to increase rather than lower reactivity (Allen and others 2002; Christian and Stoney 2006). Allen and colleagues (2002; 1991), comparing different sources of social support, found that reactivity tended to be higher in the presence of a spouse, but lower with support from pets. These authors suggest that pets serve an important role as non evaluative sources of support. Women participating in a laboratory study with a friend present had significantly higher lipid reactivity (triglycerides, total cholesterol) to the stressor than women with no support (Stoney and Finney 2000). The nature of the ties may also be a factor: individuals with a higher number of ambivalent social ties had greater SNS mediated heart rate reactivity to a laboratory stressor as a function of age (Uchino and others 2001). Finally, several investigators have focused on personality traits which may moderate the benefits of social support on cardiovascular reactivity. For example, some studies provide support for the conclusion that social support is of less benefit to individuals who are cynical or defensive
(Lepore 1995; Westmaas and Jamner 2006).
Other studies have focused on ambulatory blood pressure and heart rate, in relation to a range of social interactions, based on the prediction that changes in cardiovascular function related to everyday experiences will reveal insights not available from a laboratory setting
(Kamarck and others 1998). In a number of cases, study participants had a lower heart rate
100 and/or blood pressure in the presence of family (Holt Lunstad and others 2003; Spitzer and others 1992; Steptoe and others 2000) and significant others (Gump and others 2001; Holt
Lunstad and others 2003), and higher values when alone or in the presence of strangers
(Brondolo and others 1999; Gump and others 2001; Spitzer and others 1992). Evans and Steptoe
(2001) found that work related social support was predictive of lower heart rate, but not blood pressure, only during workdays. Linden and colleagues (1993) found that social support was related to lower ambulatory systolic blood pressure among women but not men. Finally, Steptoe and colleagues (2000) found reduced blood pressure, in the evenings after work, among individuals with children, and high social support. In contrast, Gump and colleagues (2001) found no evidence that social support mediates the relationship between spousal interactions and reduced systolic blood pressure.
Of the studies that tested affect as a mediator between social interactions and ambulatory cardiovascular outcomes, few found evidence for a relationship (Brondolo and others 1999;
Holt Lunstad and others 2003; Linden and others 1993), suggesting that perceptions of threat or comfort in a social interaction may be subcortical (Gump and others 2001). However; Holt
Lunstad and colleagues (2003) found the highest level of arousal in ambivalent relationships, specifically, those characterized by both highly positive and highly negative feelings, perhaps indicating a higher level of vigilance during these interactions. Some of the studies provided evidence for higher reactivity to social interactions among some individuals than others
(Brondolo and others 1999). In particular, Kamarck and colleagues (1998) argue that while overall effect sizes of arousal during social interactions on diastolic blood pressure were small, they accounted for a much higher percentage of daily variation for the most reactive individuals.
Finally, many of the authors did not directly test the buffering hypothesis, but Linden and
101 colleagues (1993) found no evidence that social support moderates the relationship between stress and blood pressure, as perceived stress was unrelated to physiology.
Another group of studies focused on the downstream effects of elevated blood pressure and heart rate, specifically atherosclerosis, coronary artery disease, and myocardial infarction.
Increasingly, atherosclerosis has come to be seen as an inflammatory condition, rather than one related exclusively to fat deposition in the vasculature (Mahmoudi and others 2006). The process of arterial plaque formation begins with damage to the endothelium, the smooth inner lining of blood vessels, whether through mechanical processes such as high blood pressure and heart rate, or chemical processes such as hyperglycemia (Blasi 2004). In this way, chronic sympathetic activation related to stress can play an important role in the development of disease
(Rozanski and others 1999). Inflammatory responses to the lesion promote deposition and oxidation of low density lipoprotein within the arterial wall (Mahmoudi and others 2006). As the plaque continues to develop, inflammation promotes the migration and proliferation of smooth muscle cells and lipids, and an extracellular matrix in the form of collagen and elastin, which eventually form a thick fibrous cap (Croce and Libby 2007). With continued inflammation the cap begins to break down, which can result in plaque rupture, blockage of the artery, and myocardial infarction or stroke (Croce and Libby 2007). Recent evidence indicates that patients treated for acute coronary events generally have numerous unstable plaques, highlighting the importance of systemic inflammation in disease progression. A continuing process of rupture, erosion, and repair servers to enlarge arterial plaques over time (Mahmoudi and others 2006). Atherosclerotic arteries have reduced dilator, and augmented constrictor response, but this phenomenon may be independently related to psychosocial stress rather than to the extent of arterial plaques (Williams and others 1991).
102 Much of the information on the development of atherosclerosis, as a function of social interactions or isolation, comes from research on animals (Manuck and others 1995). In two studies of cynomolgus monkeys with diet induced atherosclerosis, Williams and colleagues
(1994; 1991) found vasodilation of coronary arteries in response to acetylcholine to be restricted amongst subordinate animals, or those subject to periodic changes in social group composition.
In both cases, vascular response was related to social conditions rather than diet or degree of atherosclerosis. While this response may be independent of atherosclerosis, it may explain associations between social support and reduced risk of acute coronary events. In a series of experiments on monkeys and rabbits, Kaplan and colleagues (1991) found that the administration of β adrenergic blocking agents to significantly reduce the development of atherosclerosis among those animals subject to more social stress, demonstrating the causal importance of sympathetic activation. Endothelial cell death and platelet aggregation were significantly higher in areas subject to high shear, suggesting that hemodynamic processes are important in plaque formation (Kaplan and others 1991).
Developments in imaging technology have provided the ability to observe the degree and progression of arterial plaques in hospital and laboratory settings, usually among patients seeking treatment for associated complications. A number of studies provide evidence of a negative relationship between atherosclerosis and social integration (Knox and others 2000; Kop and others 2005; Orth Gomer and others 1998). Orth Gomer and colleagues (1993), over a six year period, found that social integration was a more important predictor of myocardial infarction among middle aged men than social support. Another study found no relationship between carotid artery atherosclerosis and social support, once potential confounders were entered into the model (Knox and others 2000). In a one year follow up of women hospitalized for
103 myocardial infarction, depression was predictive of mortality, but only among those with low social support (Frasure Smith and others 2000). Several studies have found a relationship between social support and coronary atherosclerosis: two longitudinal studies demonstrate a negative relationship between emotional social support and progression of atherosclerosis
(Angerer and others 2000; Wang and others 2005), while another found the extent of coronary atherosclerosis, was lower with higher instrumental but not emotional support. Angerer and colleagues (2000) suggest that social support is important for slowing the progression of disease, while social integration is more important as a determinant of cardiovascular mortality. While most of these studies focused on disease severity and progression, one project found higher social support to be predictive of pursuing more aggressive treatment options for peripheral arterial disease, suggesting another avenue through which social support might influence mortality (Aquarius and others 2006).
Social support and immune function
The immune system, the body’s defense against tissue infection and injury, is complex, and involves a large number of cell types with numerous different functions. For this reason,
Uchino (2006) argues that immune parameters should be interpreted with caution. I will begin with an overview of immunity, focusing on those parameters that have been investigated with regard to social support. Though I will briefly discuss the evidence related to C reactive protein
(CRP), and antibodies to the Epstein Bar Virus (EBV), I will discuss these immune parameters in more detail in later chapters.
After chemical and physical barriers such as skin, pathogens encounter non specific defenses, which target any injury or foreign cells. Included in this response are fever, which increases the metabolic rate of cells, speeding both defense and repair mechanisms.
104 Inflammation, characterize by warmth, redness, swelling and pain, begins when tissue injury promotes the release of histamine from mast cells. Histamine, in turn, promotes vasodilation, releasing phagocytes into the interstitial fluid, along with clotting proteins to wall off the affected area, and oxygen and nutrients to begin the process of tissue repair. Some important components of non specific defenses include natural killer (NK) cells, which chemically attack both cancerous and virally infected cells, and C reactive protein (CRP), an important component of systemic responses to local inflammation (Fleck 1989). CRP, part of the acute phase response, is released by the liver and rises dramatically in concentration within a day or two of initial infection (Mortensen 1994). Evidence indicates that CRP plays an important role in recognizing and destroying bacteria, among other critical functions (Mortensen 1994), but is probably most well known as a marker of inflammation associated with cardiovascular disease
(Verma and others 2007).
The final line of defense against pathogens is specific immunity, which targets particular antigens, and has the ability to remember and respond more quickly to any pathogens previously encountered, in any part of the body. There are two overall types of immunity: antibody mediated, or humoral immunity, which targets extracellular parasites and microorganisms and; cell mediated immunity, which targets both virally infected and cancerous cells (Dong and
Flavell 2001). These forms of immunity are mutually inhibitory (Elenkov 2004).
Humoral immunity relies on the action of B cells, so named because they develop in the bone marrow. B cells have unique surface receptors that allow them to bind to specific pathogens. They remain, inactive, in the tonsils, spleen, and lymph nodes until they encounter an antigen that binds to their surface receptor. Those B cells that are stimulated in this fashion produce numerous clones with identical surface receptors, which move from the lymphatic
105 system into the plasma. These clones produce antibodies, which bind to the antigen, and mark them for other components of the immune system to destroy. Memory cells are B cells with a previously activated surface receptor, which remain inactive, but can stimulate a more rapid antibody response when the organism encounters the appropriate antigen. There are five classes of antibody, or immunoglobulin (Ig), each with unique functions, but I will focus here on two.
The most common of these is IgG, found in blood, lymph, the digestive tract, and interstitial fluid, which functions to activate the complement system (plasma proteins that assist in immune defense), and neutralize toxins. IgA, present on mucous membranes, is an important first line of defense against pathogens as they attempt to enter the body via these surfaces.
Cell mediated immunity involves the action of T cells, which are produced in bone marrow, but mature in the thymus. In contrast to B cells, T cells cannot recognize antigens in their entirety, but only their fragments. Antigen presenting cells (APC), including B cells and macrophages, partially digest antigens and present the relevant portions for recognition by T cells, which then destroy the cells carrying the pathogen. Immature T cells develop in the thymus, stimulated by the hormone thymosin, to carry surface receptors uniquely able to recognize one of a nearly infinite number of potential antigen fragments. The function of T cells is determined by the possession of one of two surface proteins, either CD4 or CD8 receptors.
T cells displaying CD4 receptors differentiate into helper T cells (Th) when they encounter an APC displaying a fragment that matches their surface receptor, and produce numerous clones. Th cells secrete cytokines, or immunological signaling molecules, which stimulate phagocytes, NK cells, and cytotoxic T cells, and enhance the inflammatory response.
An important branch of cytokines are interleukins (IL), which stimulate the development of other immune cells. For example, IL 6 stimulates growth and differentiation of both B and T cells,
106 and activates hematopoietic cells in the liver, leading to the production of CRP and other acute phase proteins (Besedovsky and del Rey 1996). Th cells further differentiate into two subsets upon activation (Mosmann and Coffman 1989). Th1 cells release cytokines which mediate cellular immunity (eg. interferon gamma (IFN), and tumor necrosis factor beta (TNF)), while cytokines produced by Th2 cells regulate humoral immunity (e.g. IL 6) (Dong and Flavell 2001;
Elenkov 2004).
Cytotoxic, or killer T cells, displaying CD8 receptors, undergo a similar process of stimulation and proliferation upon encountering appropriate APCs, but directly attack foreign cells by injecting them with the toxin perforin. Upon activation, cytotoxic T cells move through the circulatory and lymphatic systems in search of their target antigen fragments, and concentrate at the site of an infection or tumor. These cells also secrete a cytokine known as tumor necrosis factor (TNF), which plays an important role in fighting cancerous cells.
A commonly used indicator of cell mediated immunity is the concentration of antibodies to the Epstein – Barr virus. Most people are infected with this virus by the time they reach childhood, and it persists in a latent state throughout their lifespan (Glaser 2005; Kiecolt Glaser and others 2002b). Individuals with appropriately responsive cell mediated immunity are able to keep the virus in check after primary infection, but a stress related shift toward humoral immunity allows the virus to multiply and results in increased viral antibodies (Kiecolt Glaser and others 2002a). A wide body of evidence supports the use of antibodies to the Epstein Barr virus as a particularly sensitive indicator of psychosocial stress (Glaser 2005).
There are numerous avenues of bidirectional communication between the nervous, endocrine, and immune systems, but this field of study is in its relatively early stages
(Besedovsky and del Rey 1996; Thayer and Sternberg 2006). The sympathetic nervous system
107 innervates the thymus, bone marrow, spleen, and lymph nodes, and immune cells have hormonal receptors (Maier and others 1994). Similarly, cytokines convey information to the endocrine and nervous systems. For example, there are receptors for IL 6 in both the brain and the pituitary gland (Besedovsky and del Rey 1996). Recent evidence suggests that glucocorticoids and catecholamines suppress Th1 mediated cellular immunity, relative to Th2 mediated humoral immunity (Elenkov 2004). This has important consequences in terms of increasing systemic inflammation, and compromising the body’s ability to fight infections such as EBV, HIV, tuberculosis, and H. pylori (Elenkov 2004) . There may, however; be different responses to acute and chronic stress. For example, laboratory stressors increase the number of circulating cytotoxic T cells, while long term stressors have the opposite effect (Herbert and Cohen 1993).
Finally, there is also some evidence that oxytocin inhibits the release of IL 6 from the neurointermediate lobe of the pituitary gland (Besedovsky and del Rey 1996).
Animal studies demonstrate that isolation negatively affects immune function, and that early experience can have life long consequences. Rhesus monkeys isolated in early life, then reintroduced to a captive colony, had a decreased ratio of helper to suppressor T cells, and an increase in the number and activity of NK cells, and isolated males, in particular, had a significantly higher risk of early death. Similarly, early life isolation influences the immune response to pathogens. Piglets isolated two hours a day demonstrated reduced TNF α response to endotoxin challenge, along with more intense symptoms of illness, such as vomiting and shivering (Tuchscherer and others 2006). Socially isolated mice injected with cancerous cells were far more likely to develop tumor colonies in the liver, and were significantly less responsive to chemotherapy than mice in a social setting (Wu and others 2001).
108 A number of authors have tested the importance of social support and immune function with relation to certain specific stressors such as college exams and marriage breakup. Among military trainees at West Point, final exams, in contrast to basic training, were related to an increase in EBV titers (Glaser and others 1999). The authors suggest that the fundamentally social nature of basic training, as opposed to the solitary nature of exam stress, may explain the lack of stress response to such an intense experience. NK cytotoxicity was negatively related to exam stress among healthy and asthmatic adolescents, but this relationship was somewhat attenuated among those with higher social support (Kang and others 1998). Secretory IgA was reduced among undergraduates taking an exam, but social support was positively related to this antibody only during a low stress period, demonstrating an effect contrary to stress buffering
(Jemmott and Magloire 1988). T cell proliferation in response to EBV polypetides was lower for medical students during exam time, and among those who reported higher support seeking behavior (Glaser and others 1993). Kiecolt Glaser (1987a) found that marital quality was related to a number of immune parameters including T cell proliferation, NK cells, and EBV antibody titers, and that immunological profiles were worse among those recently separated from their spouse, particularly if their attachment was high. Finally, social support predicted higher T cell proliferation to one mitogen, and NK cell lysis, (Baron and others 1990) and improved response to Hepatitis B vaccination (Glaser and others 1992), among spouses of cancer patients.
Loneliness and total network size were related to antibody production in response to vaccination against a strain of the flu (Pressman and others 2005). Though loneliness was also related to cortisol, it did not mediate the relationship between loneliness and antibody response.
In addition, there was an interaction such that those individuals scoring high on loneliness and low on social network demonstrated the worst response. Marsland and colleagues (2007) found
109 that IL 8 was positively related to stress, and negatively related to social support, with little evidence of buffering, but the same was not true of IL 6. In a study of middle aged adults, IL 6 was negatively related to social integration among men, but there was no such relationship with
CRP (Loucks and others 2006). Finally, Miyazaki and colleagues (2005) found support for both the buffering and direct hypotheses. Social support was directly predictive of NK cell counts,
IL 4, and Th1/Th2 balance, while it buffered the effects of perceived stress on T cell counts and in vitro INF y production.
Several studies of social relationships and immune function have focused on cancer patients, generally women with ovarian or breast cancer. Reasons for this include that the immune system is critical in fighting cancer, and having cancer represents an independent source of stress. Cancer patients with higher IL 6 and reduced NK activity tend to have poorer outcomes, and both of these immune parameters are predicted by stress (Costanzo and others
2005; Lutgendorf and others 2000; Von Ah and others 2007). Von ah and colleagues (2007) found that perceived stress was negatively related to NK cell activity among post operative breast cancer patients, while social support had no buffering effect. In contrast, Lutgendorf and colleagues (Lutgendorf and others 2000) found that IL 6 was significantly higher among gynecologic cancer patients than healthy controls, and that instrumental support seeking was negatively related to this cytokine only among the patients. Social attachment was negatively related to IL 6 (Costanzo and others 2005), and social support was positively related to NK activity/cytotoxicity among ovarian cancer patients, both in peripheral blood and in the vicinity of the tumor (Levy and others 1990; Lutgendorf and others 2005).
There is some evidence that increased inflammation during pregnancy confers a higher risk of complications such as gestational diabetes, preeclampsia, and pre term delivery
110 (Coussons Read and others 2007; Coussons Read and others 2005), perhaps related to physiological responses to infection (McGregor and others 2000). As discussed in chapter two, similar inflammatory responses result from the stress of high altitude hypoxia (Coussons Read and others 2002). Coussons Read and colleagues (2007; 2005) found perceived stress to be strongly related to the pro inflammatory cytokines IL 6, IL 10 and TNFα, and to CRP, among pregnant women. Social support explained a much smaller percentage of the variation in each of these outcomes, and did not moderate their relationship with stress.
Another population of interest to social epidemiologists has been older adults, as aging brings about a decline in immune function (McDade 2003), and variation in immune function has important implications for diseases such as Alzheimer’s and osteoporosis (Friedman and others 2005). NK cells may be mediated by stress related changes in cytokines, with important implications for the development of cancer (Esterling and others 1996). Esterling and colleagues
(1994b; 1996) found that NK cell activity is reduced among both bereaved and continuing caregivers, relative to controls, and that social support is significantly predicative of better NK activity in this group. Though IL 6 tends to be higher among older adults, high sleep quality, and more positive social interactions protect against age related increases (Friedman and others
2005). CRP, a downstream product of IL 6, is negatively related to social integration for men over 60, but not women or younger men (Ford and others 2006). Finally, fibrinogen, a cytokine which promotes clotting and thrombosis, was negatively associated with social integration among men, but not women, enrolled in the McArthur successful aging study (Loucks and others
2005).
As discussed by a number of these authors, the effect sizes of stress or social support on immune parameters vary, and their functional significance is often unclear (e.g. Kiecolt Glaser
111 and others 1987a). As discussed above, immune function has implications for response to mitogens and vaccination, suggesting that psychosocial modulation of immunity has implications in terms of host resistance to infection. In particular, cell mediated immunity is the primary means of fighting viral infection of host cells, so a stress related shift toward humoral immunity is likely to impair this process. Cohen and colleagues (1997; 2003) found that social relationships predict infection after experimental exposure to a cold virus. In one case, those individuals who scored higher on sociability were less likely to become infected (Cohen and others 2003). Though sociability was related to having a larger social network and greater access to social support, these did not mediate the relationship between sociability and infection. In another case, these authors found network diversity, rather than the number of network members, to be related to risk of infection, and suggest that variation in host resistance may explain some of the relationship between social integration and all cause mortality (Cohen and others 1997).
Though some studies have found no relationship between progression of HIV infection and social support, Uchino and colleagues (1996) suggest this may relate to the difficulty of controlling for a number of confounders, including disease stage. Some studies indicated a relationship between emotional social support and CD4+ counts during the latter stages of the disease (Theorell and others 1995; Uchino and others 1996).
Social support and nutritional status
Finally, very few studies address the relationship between social support and nutritional status. Social support, particularly instrumental support, may influence economic welfare and improve dietary intake. Social support is an important predictor of food security for the elderly
(Locher and others 2005), and, potentially, for households headed by women (Lemke and others
2003). Hadley and colleagues (2007), working in Tanzania, found that perceived social support
112 was related to greater food security only among wealthier households, suggesting either that these households have more connections, or are better able to draw upon them. Affectionate support for mothers in Brazil was related to higher weight for height and weight for age Z scores for children, while lack of material support from their network was related to lower weight for height Z score (Surkan and others 2007). Contrary to their prediction, having access to social support when dealing with conflict with a spouse, relative or friend was negatively related to child stature, as characterized by height for age z score (Surkan and others 2007).
Based on the association of high waist to hip circumference ratio (WHR) with diabetes and coronary heart disease, and preliminary evidence that this association may be driven by stress related dysregulation of the HPA and sympathetic axes, Ravaja and colleagues (1998) explored the influence of social support on waist to hip circumference ratio (WHR), a correlate of diabetes and coronary heart disease risks. , Using a prospective study design, the authors found that improvements in emotional social support were associated with an increase in WHR among adolescent and young adult males, but only marginally so among females
Although some of the findings are negative or contradictory, the balance of evidence supports the conclusion that social relationships, and social support, have important implications for a broad range of health outcomes. Much of the evidence is consistent with the hypothesis that social support influences health by buffering HPA and Sympathetic adrenomedullary activation, and their downstream effects. However; a number of studies found evidence for direct effects of social support on these axes, and many more did not test the stress buffering model. There is also considerable evidence that social isolation is an independent source of stress, which can activate these same systems. The idea of allostasis as shifting set points in response to average levels of demand is an attractive one. It is possible that a long term rise in
113 blood pressure in response to repeated elevations enhances adaptive flexibility, and is associated with relatively less wear and tear than would otherwise be the case. This seems unlikely given the broad physiological damage associated with upregulation of these systems. The evidence is more consistent with the idea of allostatic load (McEwen 2007), and a reduction in adaptive flexibility as increasing demands are put upon the system.
Hypothalamus: Stress Oxytocin Corticotropin releasing hormone Social Central nervous system support Brain Spinal cord Pituitary gland, anterior lobe: Corticotropin (ACTH): glucocorticoids (cortisol) Peripheral nervous Pituitary gland, posterior lobe: system Food Oxytocin Autonomic division: security
Nutritional status: Body composition Adrenal gland, medulla: Epinephrine Immune function: Norepinephrine Inflammation (CRP) Immune function Adrenal gland, cortex: (EBV) Cortisol
Figure 3.2. Physiological pathways linking social support, shown here as a buffer of stress, and the outcomes of this study: body composition, inflammation, and immune function (the colors highlight different pathways through which social support might relate to nutritional status).
Figure 3.2 presents a simplified model of the theorized relationship between social support and the outcomes used in this study. Based on the above discussion, I predict that social
114 support will be negatively related to inflammation, and positively related to immune function. I also predict that social support will be positively related to women’s levels of body fat, because it will protect against food insecurity, however; social support may be negatively related to body fat if it protects against the effects of stress on visceral adiposity.
In the preceding sections I focused on the physiological pathways associated with stress and social support. One potential reason for the diversity of findings reviewed above is that there is little consistency in measuring social constructs, and a tendency to focus on a narrow range of potential benefits associated with social relationships, in particular, emotional social support. In the next section I will examine evidence for the importance of different types of social support for health, along with issues in the measurement of this construct.
Social support, measurement and culture
There are several conclusions that can be generalized from studies of social support and health. In their meta analysis, Uchino and colleagues (1996) found that the greatest effect sizes are associated with emotional support, and support from family, though emotional support is also the dimension most often investigated. In addition, as discussed above, there is evidence for both direct and buffering effects of social support on physiology (Uchino and others 1996). In a small number of the studies reviewed above, instrumental and appraisal support were more strongly related to health than emotional support. Cohen and Wills (1985) have suggested that emotional support serves as a general buffer of psychosocial stress, while other forms are important only when they are relevant to the stressor being experienced. Based on this analysis, we may predict that the relevance of instrumental support varies with macroeconomic conditions and with individual or household resources. There is some evidence that the influence of emotional support (Dressler and others 1997) and instrumental support (Hadley and others 2007;
115 Strogatz and James 1986) vary based on socioeconomic status or class. Instrumental support may buffer stress associated with economic circumstances or social status, but to the extent that it actually improves economic circumstances, it may act directly to improve financial coping mechanisms (Berkman and others 2000; Thoits 1995).
There are several methodological difficulties associated with the measurement of social support. Researchers have developed a large number of well validated instruments for measuring social support, but they tend to be relatively narrowly focused, and each one examines different functional aspects of social relationships, while, in contrast, non standardized measures rarely report psychometric data (Winemiller and others 1993). In each case, comparison between studies is of limited value. Investigators also tend to limit their explanation of their measurement strategy, making interpretation of the results somewhat more difficult (Winemiller and others 1993). Also limiting comparability is the fact that these instruments have been developed for a number of different purposes. The more detailed the questionnaire, the more information is provided on functional and qualitative aspects of social relationships, but this must be balanced by practicality in research settings (Orth Gomer and Unden 1987). Many of these instruments have been designed for purposes such as identifying individuals with low social support as a risk factor associated with various health conditions, and so provide limited information for research purposes. Finally, these instruments often ask people to report their access to social support with relation to specific needs or problems, for example, having someone who can drive them to the airport (Winemiller and others 1993). These questions are obviously highly context specific, and would potentially underestimate social support among individuals of low SES.
116 There are also several potential aspects of supportive social relationships which have thus far received limited attention. First, very few studies consider the degree to which socially supportive relationships are reciprocated, and the potential importance of this for biological variation (Jacobson 1987; Winemiller and others 1993). For example, individuals who provide more support than they receive may be uniquely stressed. In addition, some of the studies reviewed above suggest that some aspects of social relationships, even ones perceived as supportive, may have negative consequences for health, but few investigators have explicitly examined this issue (Dressler and others 1986; Jacobson 1987; Seeman and McEwen 1996;
Uchino and others 1996).
Another consideration is that, though many studies suggest that the perception of support is more important than absolute levels of support for buffering the stress reaction, few authors have examined differences in received and perceived support (Jacobson 1987; Winemiller and others 1993). In particular, Jacobson (1987) suggests that attempting to draw on support that turns out to be unavailable, may lead to a greater stress reaction than not having support to begin with. There are, however; a number of potential problems with assessing received support, including the potential to confound support need and availability (Cohen and Wills 1985). The timing of support provision may also be critical in determining whether it is perceived in a positive or negative light (Jacobson 1987).
There is some evidence of gender differences in the influence of social support on physiology (Cohen 2004; Dressler and others 1986; Seeman and McEwen 1996; Seeman and
Syme 1987; Taylor and others 2000; Uchino and others 1996). Some investigators have suggested potential differences in hormonal mediators for men and women. For example, Taylor and colleagues (2000) suggest that females tend more towards a pattern of “tend and befriend”
117 than fight or flight, using investment in social relationships, mediated by oxytocin, to protect themselves and offspring. Others have emphasized the importance of cultural norms and expectations regarding sociality. For example, women may experience a greater cost in social relationships, as they are often expected to serve as caregivers to others (Cohen 2004; Dressler and others 1986; Seeman and McEwen 1996). Others have suggested that involvement in large family networks may place more constraints on behavior for women than men, influencing the benefits of drawing on various kinds of support (Brown and Lapuyade 2001).
Finally, and related to all of these issues, is the importance of culture as an influence of the relationship between social support and health. This has been recognized as important by a large number of authors, but has received relatively limited attention to date (Berkman and others 2000; Dressler and others 1997; Dressler and Bindon 2000; Jacobson 1987; Taylor and others 2004; Taylor and others 2007; Thoits 1995; Uchino and others 1996). Culture specifies the norms associated with various social relationships, and under what general circumstances these relationships are appropriately drawn upon, which can influence the costs and benefits of doing so (Dressler and Bindon 2000; Dressler and others 1986; Dressler and Santos 2000). In particular, cultural classifications of kinship and fictive kinship serve, in part, to define expectations associated with specific categories of relationship, including expectations of support
(Jacobson 1987). The vast majority of studies on social support and health have been conducted in a western setting, suggesting that much of current literature takes for granted western cultural norms.
Among psychologists, Taylor and colleagues (2004; 2007) are unique in their focus on the cultural dimensions of social support. According to these authors, Asians, and Asian
Americans report less reliance on social support, and demonstrate less physiological benefit from
118 supportive relationships. These authors suggest that the individualistic nature of western culture makes social support more appropriate in this setting, but Asians tend to subordinate their needs to those of their family and social network. These authors used focus groups of Asian
Americans to discuss specific reasons for these broad patterns, and then used a questionnaire to examine them in a larger sample. There were many explanations suggested by study participants, but the most common was that people were reluctant to draw on support because of concern for damaging relationships, and for fear of criticism (Taylor and others 2004). In a subsequent study, these authors found that Asian Americans benefit more from implicit social support (focusing on valued social groups) than explicit social support (Taylor and others 2007).
One potential drawback of these studies is that the investigators made broad generalizations about Asians and Asian Americans from a number of different countries.
Several anthropologists have relied on qualitative ethnographic methods and anthropological literature to develop an understanding of cultural norms of social support.
Jacobson (1987) advocates thick description, or a meaning centered approach, to understanding social support, based on participant observation and in depth interviews. In a qualitative study of the elderly, he 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. In addition individuals perceive negatively support which they interpret to challenge their independence. He does not, however; address the applicability of these methods in quantitative studies. In their study of Samoans living in California, Janes and Pawson (1986) emphasize the critical importance of extended family ties, and the potential cost of having a large number of family obligations. These authors found that having a larger supportive social network was associated with lower blood pressure, and suggest that social support buffers the
119 relationship between obesity and blood pressure. Finally, working in a community in central
Mexico, Dressler and colleagues (1986) found gender differences in social support which they suggest are based on cultural norms. In this community 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 benefited 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; seeking support outside this network violates these cultural norms and causes stress.
More recently, Dressler and colleagues (1997; Dressler and Bindon 2000) used consensus analysis to examine the cultural dimensions of social support. Working 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. Among Brazilians, consensus analysis indicated that, for most problems, it is culturally appropriate to call on family, friends, neighbors, then compadres (co parents) for support (Dressler and others 1997). Cultural consonance, or the degree to which 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. Among African Americans, the cultural model indicated the unique importance of kin support relative to all other sources (Dressler and Bindon
2000). These authors found an interaction between kin support and cultural consonance in lifestyle such that there was a linear negative relationship between lifestyle consonance and blood pressure for those with high kin support, while the relationship was curvilinear for those
120 with low kin support. That is, individuals with low and high cultural consonance in lifestyle had higher blood pressure than those in the middle of the range. These authors suggest that achieving high social status outside of the context of supportive family is uniquely stressful in this community.
These studies suggest that both macroeconomic and cultural context can play an important role in structuring the costs and benefits of social support and social relationships.
Anthropologists are in a unique position to link these larger scale forces to individual social relationships and biological outcomes. Specific tools that can be used to address this goal include immersion in the regional and local ethnographic, historic, and political economic literature, ethnographic methods including interviews and participant observation, and more structured ethnographic methods such as consensus analysis; the goal being to link cultural and political economic context to individual measures of social support and health. In the following section, I will discuss social relationships in the Andes, focusing on how these might be expected to influence the costs and benefits of social support.
Kinship, fictive kinship and gender relations in the Andes
Patterns of kinship vary across the Andes, and are often dependent on local economic circumstances. In rural areas, it is generally more common for recently married or common law couples to live with the husband’s parents until they are able to become economically independent, however; this depends on considerations such as the relative wealth and access to land of the partners (Bindon and Vitzthum 2002; Buechler and Buechler 1971; Mayer 2002;
Paerregaard 1997; Van Vleet 2002). Couples build their independence through a process of fulfilling obligations to their kin and fictive kin, and strengthening economic relationships
(Weismantel 1995; Weismantel 1988). For women, this often involves economic service and the
121 demonstration of respect to their husband’s family, particularly their mother in law (Van Vleet
2002). In many parts of the Andes, inheritance is parallel, so that husbands inherit land from their father, and women from their mothers, with both partners retaining control of their own resources (Collins 1986); while in other areas, perhaps based on Spanish social norms, inheritance is through the father, and biased towards sons (Buechler and Buechler 1971;
McEwen 1975).
A number of Andean ethnographers have questioned the universal applicability of the household or biological family as a fundamental unit of analysis. Many feminist scholars have argued for a more detailed consideration of social dynamics within the household, rather than taking it as an irreducible unit of production and reproduction, whether this involves a discussion of conflict within families, or cultural norms in the Andes which define kinship based on nourishment and mutual obligation rather than biological distance (Nash 1993a; Van Vleet 2002;
Weismantel 1995). Collins (1986), in particular, argues that the appearance of the household as the most important locus of economic production is recent, and related explicitly to capitalist development. For example, among rural Aymara, women and men each 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). She argues that an increasing emphasis on the nuclear family is evidence of a loss of social resources; while engagement in wage labor increases the difficulty of creating these ties, inadequate earnings make them critical for continued subsistence production (Collins 1986). This is in keeping with other authors who argue that neoliberalism is associated with a weakening of supportive networks, and household retrenchment (Gledhill
2005).
122 Compadrazgo is an important means of developing extra household ties in the Andes.
This is a Spanish term for a custom from Europe, 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) .
This form of fictive kinship is common to Latin America, but the benefits and expectations associated with these relationships vary from one location to another based on local cultural practices, and political and economic circumstances (Mintz and Wolf 1950).
Anthropologists have had a longstanding interest in the degree to which fictive kinship benefits the parties involved (Crandon Malamud 1993; Sayres 1956), varies in expression based on class (van Den Berghe and van Den Berghe 1966), extends social networks or strengthens existing ties (Deshon 1963; Foster 1969), and varies in form between rural and urban settings
(Carlos 1973; Kemper 1982; Middleton 1975). This diversity is reflected in ethnographic work in the rural Andes, demonstrating 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 1993b; Paulson 2006; Weismantel 1988) and as a means for establishing exploitive economic relationships based on race and class, for example, between meztizos and campesinos (Albro 2000a; Mayer 2002; Nash 1993b; Paerregaard 1997;
Seligmann 2004; Weismantel 1988).
Previous ethnographic work suggests that economic cooperation can be one of the main benefits of ritual kinship in rural areas, and that compadrazgo forms the primary means of extending ayni , or reciprocal work exchange, beyond kin (Allen 1988; Collins 1986).
123 Weismantel (1988) has argued that access to business partners, loans and agricultural labor are forms of wealth built up through a lifetime of compadrazgo ties, but the landholdings associated with this strategy are increasingly unavailable to younger generations. As with larger kinship networks, involvement in wage labor weakens fictive kinship as a source of labor and economic collaboration (Collins 1986; Leatherman 1994). However; other authors suggest that these relationships can be an important means of strengthening pre existing social relationships, and structuring social life (Buechler and Buechler 1971).
Research on compadrazgo within cities has been relatively limited, but indicates that, though it may operate in subtly different ways, ritual kinship can continue to play an important role in urban life (Carlos 1973; Kemper 1982; Middleton 1975). In addition, rather than being a separate phenomenon, it can represent a strategy for establishing rural to urban economic ties, for example between wholesalers or market vendors and local producers (Crandon Malamud 1993;
Seligmann 2004). In El Alto, fictive kinship may be an important source of support for newer migrants (Buechler and Buechler 1971), and is involved in structuring economic exchange related to participation in, and organization of festivals and parades, important aspects of cultural life in this city (Lazar 2008). In general terms, cultural anthropologists have long been interested in the degree to which economic cooperation operates in marginal urban areas. While some researchers have found that reliance on supportive networks is an important survival strategy in marginal urban settings (Halebsky 1995; Lomintz 1978), other researchers have found that extreme poverty limits the benefits associated with reciprocal exchange (Winter 1991). The bulk of anthropological work on fictive kinship took place in the 70s and 80s, and has more recently dropped out of favor, but as discussed by Jacobson (1987), this may have profound implications for cultural constructions of social support, particularly instrumental or economic social support.
124 Finally, a number of authors have proposed that Andeans have a tendency toward more egalitarian gender norms than seen in many other regions, including urban areas in the Andes
(Hamilton 1998). Specifically, gender complementarity, or the idea that males and females have separate, well defined, and mutually important gendered economic roles, assigns values to both genders (Allen 1988; Harris 1978; Harris 1980). While men tend to have more power in the public sphere, and women in the domestic sphere, holding local political or leadership roles, while they perform different aspects of this process, is often something done by couples rather than individuals (Buechler and Buechler 1971). In Aymara, the term chachawarmi , where chacha means male and warmi means female, refers to a married couple as a complementary and irreducible unit (Harris 1978). In practice, however; complementarity is an ideal, and an organizing principle, rather than an absolute value. This understanding of gender roles does not preclude high rates of domestic violence directed against women, and exclusion of women from the political process apart from their role in feeding men (Allen 1988; Harris 1978; Harris 1980;
Van Vleet 2002).
Gender norms may also be somewhat more fluid among indigenous Andeans than is true of the dominant criollo culture. Harris (1980) argues that while values of strength and courage are associated with maleness, and weakness and fear associated with femaleness, men and women can be assigned either role based on their behavior. Though there are norms regarding how individuals take on opposing gender identities, for example in the labor they perform, men and women are thought to have characteristics of both (Allen 1988). Fundamentally tied up with ethnicity, are gendered stereotypes of cholas. Cholo , a highly contested term, generally refers to urban mestizos who maintain aspects of indigenous identity, while distinguishing themselves both from rural campesinos and criollo elites; for women, primarily in their manner of dress
125 (Stephenson 1999; Weismantel 2001). Cholas are often associated with maleness: for their ability to move between rural/urban, public/private spaces, through code switching, dress, and economic activity (Albro 2000b; Paulson and Calla 2000; Seligmann 1993; Weismantel 2001); for their relative power in dealing with criollos during market exchanges, and with local producers (Seligmann 1993), and; for their bawdy humor, self assuredness, and economic independence (Albro 2000b; Seligmann 1993).
Donning the pollera, the layered skirts of the chola , represents a form of resistance and cultural contestation in a patriarchal and racist society (Stephenson 1999; Weismantel 2001).
Criollos from La Paz describe market cholas as predatory, dishonest, and wealthy, and, similar to other regions, often use explicitly racist language in their market encounters with these women
(Seligmann 1993). Through their dress, and occupation of the male public sphere, Cholas reject dominant cultural ideas of femininity, and ethnic integration (Stephenson 1999; Weismantel
2001). Women can perform different identities by switching back and forth from the pollera to western clothing depending on the circumstances, or at different periods in their life (Buechler and Buechler 1996; Paulson and Calla 2000).
As discussed in the two previous chapters, rural to urban migration in many regions of the global south is the result of the negative effects of neoliberal globalization on small scale farmers. Increasing involvement in wage labor, due in part to the increasing difficulties of agricultural subsistence, further limits agricultural investment, as well as investment in the social relationships which make agriculture feasible in the unproductive and highly variable environment of the altiplano (Carey 1990a; Collins 1986; Collins 1988; Leatherman 1996;
Leatherman 1998b). At the same time, wage labor fails to provide adequate income for household production and reproduction. In this context, migration to the city and involvement in
126 the informal economy represents a rational strategy for households and individuals. In this section, I have briefly discusses some of the changing social norms associated with shifting rural economies, however; relatively little is known about the importance of kin, fictive kin, and extra household support in urban areas.
Conclusion
I began this chapter by reviewing evidence for the relationship between social support and biology, arguing that social relationships are critical to human health: social isolation is a risk factor for a number of pathological conditions, and supportive relationships appear to mitigate other challenges to physiological function. The majority of these studies were carried out in western settings by psychologists and epidemiologists, and so have focused most closely on elucidating physiological processes, and examining psychological mediators connecting social support and health outcomes. A number of authors, however; have argued for an increasing focus on the macro social and political economic currents that influence small scale social interactions (Berkman and others 2000; Thoits 1995; Uchino 2006). In the latter part of this chapter I reviewed a smaller number of studies, highly supportive of the conclusion that class and the cultural construction of social ties have a profound influence on the relationship between social support and health. In El Alto, political economic effects on household relations and fictive kinship ties may have profound effects on household production and reproduction, and have the potential either to enhance or constrain household survival strategies with the retreat of state services. Starting in the next chapter, I will explore these issues using data collected in El Alto, Bolivia.
127
Chapter 4: Study design and methods
Introduction
As discussed in the introduction, I designed this study to assess the importance of instrumental and emotional support among women living in the marginal community of El Alto,
Bolivia. In order to accomplish this goal, I collected cross sectional data on social support, economic and household variables, and health, along with qualitative data based on interviews and participant observation. In this chapter, I will outline the methods and procedures I used to obtain these data and test the following objectives.
1) To examine the relationship between household composition, poverty, and health among women in El Alto, Bolivia.
2) To examine how culturally relevant social support influences multiple dimensions of women’s health.
3) To examine how household composition and socioeconomic status influence the relationship between social support and health.
Sampling procedures
Participants in this study were drawn from a women’s knitting cooperative, located in El
Alto. All of the women live full or part time in El Alto, in a range of different neighborhoods, or
Viacha, a town more or less contiguous with El Alto. Women from the knitting cooperative complete most of their work at home, but spend one afternoon of the week at the knitting workshop to drop off completed items and pick up supplies and orders. All of the data collection procedures were carried out with the assistance of an experienced local research assistant. My
128 research assistant and I completed all of the interviews and biological data collection in a separate room in the workshop, on women’s normal workdays. Human subjects of approval for this study was granted by the Northwestern University Institutional Review Board, while a local review was conducted by Doctora Lieselotte de Barragan at Hosptial San Gabrial , in La Paz.
During the first phase of the study unstructured interviews were collected and tape recorded from six women, to begin constructing questions for the structured interviews. Specific goals for these interviews included gaining a qualitative understanding local idiomatic terms relevant to the study and understanding what kinds of questions would be appropriate for assessing constructs like socioeconomic status and social support. Digital recordings were made of these interviews, unless the participants preferred otherwise, in which case my research assistant and I took separate notes. During the second phase, larger semi structured interviews were collected from a group of twenty nine women to further assess the utility of possible questions, and obtain broader information than would be collected in the final surveys. In each of the first sets of interviews, we used convenience sampling methods, based on who was around at the time, volunteered to participate, and provided informed consent.
In the final phase, we conducted structured interviews and biological data collection among as many women as chose to participate, and provided informed consent. 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 knitting cooperative, we contributed food to the daily meal shared by women in the workshop. In addition, women were given general information about their health, and iron supplements if they were found to be anemic.
129 Preparations: February mid April Obtaining local institutional support, obtaining permission from knitting cooperative, completing paperwork for one year work visa Phase 1: (N=6) Training of research assistant, unstructured interviews mid April May Phase 2: (N=29) Semi structured interviews Phase 3: (N=91) June – August Information sessions on biological data collection for each knitting group Structured ethnographic interviews, collection of biological Data entry, participant observation September November Table 4.1. Schedule of fieldwork for February through November 2006.
Ethnographic methods
Phase 1: unstructured interviews
In depth interviews were conducted with 6 women to discuss, among other things, their social relationships, history of work and migration, household composition and economic circumstances. Responses to these interviews were used as a qualitative test of relevant questions, and to get a sense of women’s life experiences and current circumstances.
Phase 2: semi structured interviews
More structured interviews were conducted to further test the clarity and wording of questions and to assess the nature of the variation or categorization of several specific domains: a number of questions were asked about illness and treatment categories to design questions on subjective health; women were asked to list commonly consumed foods to design a food frequency questionnaire (these data will not be reported here), and; women were asked to discuss their access to, and provision of social support to assess the most common forms and sources of both emotional and instrumental support. Throughout this process, my research assistant and I conducted detailed discussions of the clarity and validity of each question, and any possible
130 questions that we might have missed. She was also able to correct my wording when I used uncommon or confusing terms, and to clear up any confusion that I had about women’s answers to our questions.
Throughout the period of my fieldwork, I also engaged in participant observation, for example, by accompanying knitting instructors on a trip to a rural village; by attending parties and social events at the workshop or women’s homes, and; by spending several weeks at a rural agricultural college interacting with students.
Phase 3: structured interviews (Appendix A) and biological data collection (Appendix B)
Independent variables:
Household Composition. Women were asked to report and describe their relationship to other individuals in the household, and the occupation of each of the members. The criteria for household headship included being classified by oneself as the household head, being primarily responsible for family decision making and being economically responsible for other household members (Staten and others 1998). For the purposes of this project, female headed households include those households in which a woman is the primary breadwinner and economic decision maker, regardless of marital status, and those households in which no male is resident. A household was considered to have dual headship when women reported that both she and her husband fulfill the criteria presented above, and share household responsibilities relatively equally. An additional category was added to include women still living at home, with or without children.
Social relationships. Women were asked whether they had compadres, padrinos , and family in El Alto or La Paz, and how often, and for what reason they return to their rural community.
131 Instrumental social support. Women were asked, in general terms, to list the individuals to whom they would turn in the event that they needed money. With regard to previous social support, they were asked if they had ever needed help with several specific problems, and whether, and from whom, they had been able to get this help. These problems included needing help finding work or a place to live, help with an illness or childcare, or a loan. Similarly, we asked the participants who they would turn to now if they had similar problems, including needing a loan, help with childcare, help with a health problem. In each case, the answers were recorded as the number of sources of support, and the order in which they would be contacted.
Finally, with regard to specific relationships, we asked the women whether they could ask members of their family, their padrinos , or compadres for money in case of an emergency.
Emotional social support. As with instrumental support, we asked participants to list people to whom they would turn when they needed emotional support, or to talk about a problem with someone they trust. Past problems for which they reported support received included problems with their husband or mother in law. Potential problems included problem with the spouse or mother in law and problems at work. They were also asked if they could turn to members of their family, their padrinos, or compadres for emotional support, or to talk about a problem with someone they trust.
Socioeconomic Status. (SES) may be related to social support, is likely to be lower among female headed households, and may be related to any of the health outcomes. All of 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. On the recommendation of my research assistant and leaders of the kitting cooperative, all of whom thought it would be impolite, I did not ask women to report their income or about their access to
132 consumer goods. Instead, we asked whether women 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 lodgings through anticrítico . This last category, by all accounts unique to
Bolivia, is something like a lease, except that the entire sum is given to the landlord up front, then returned to the tenant at the end of the term of residence, requiring renters to have access to a relatively large lump sum of money.
Control Variables :
Neighborhood of residence. We began by asking participants which area of the city they lived in, and how long they had been living in El Alto. We also asked them how many times per year they return to their rural village, and for what reason.
Subjective Health. Self reported health can be a relatively reliable indicator of health status (Rousham and others 1998) and reflects the lived experience of disease. Women were asked to report the number of illness experienced within the last two weeks; to describe the symptoms associated with this illness; whether it was present before the reporting period; what kind of treatment they sought for the illness; whether this illness impeded their work or housework, and; how many days they spent in bed as a result of the illness.
Age and reproductive history. Participants were asked to report their age in years, how many children they had born, if any had died, and if so, what was the cause. They were also asked whether they were currently pregnant or breastfeeding.
Religious Affiliation. Church attendance is an important aspect of social life in many parts of Latin America, and might be an independent source of social support or social integration.
We asked participants with which church they were affiliated, how many times per month they
133 attended services, and whether they received any moral or economic support from members of their church.
Biological measures
Nutritional Status. Anthropometric dimensions are useful for assessing both short and long term nutritional adequacy in children and adults (Frisancho 1990). Measurements of stature, weight, mid arm circumference and skin folds, were collected following Frisancho
(1990). Skinfold measurements were taken at 4 sites (triceps, biceps, subscapular, suprailiac), using Lange Calipers, to the nearest 0.5 mm. These measurements were taken three times and averaged. Arm circumference was measured in cm using a flexible measuring tape, to the nearest mm. Stature was measured in cm using a portable stadiometer, to the nearest mm. These measurements were used to compute sum of 4 skinfolds (SSF), body mass index (BMI) (kg/m 2 ), and upper arm muscle area (UMA) following Frisancho (1990). Weight in kg (nearest 100g) 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. The student PI collected all of the anthropometric measurements.
Blood spot collection. Finger prick blood spots were collected following McDade and colleagues
(2000a). One finger was cleaned with an alcohol pad, and a sterile, disposable microlancet was used to make a small puncture. The first drop of blood was wiped away using a laboratory tissue, to clean up cellular debris, and the second drop of blood was used to test hemoglobin concentration. Then up 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
Biología de Altura to be stored at 18˚C. The samples were then carried by the student PI to the
134 Laboratory for Human Biology Research at Northwestern University and stored at 30˚C prior to analysis. The samples were exposed to tropical temperatures for less than three days, within the limits necessary to maintain sample integrity for CRP analysis (McDade et al., 2004).
Stress Epstein Barr Virus. (EBV) is a herpes virus which 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 and others 2000a). A number of studies have linked elevated antibodies to EBV with stressful experiences such as examinations (Glaser and others 1999; Glaser and others 1993; Sarid and others 2001); poor quality marriage (Kiecolt Glaser and others 1987a; Kiecolt Glaser and others 1988), caring for parents with Alzheimer’s (Kiecolt Glaser and others 1987b), space flight (Stowe and others
2001); and isolation during Antarctic expedition (Mehta and others 2000). EBV antibodies are also elevated among individuals demonstrating high reactivity to laboratory stress (Cacioppo and others 2002), and among students after verbal disclosure of stressful events (Esterling and others
1994a). In addition, this marker has been successfully used in an anthropological study designed to test the levels of stress associated with modernization and status incongruity in Samoa
(McDade 2001; McDade 2002; McDade and others 2000b). Finally, elevated EBV appears to be associated with inflammation in atherosclerotic lesions (de Boer and others 2006; Waldman and others 2008).
While most healthy individuals are able to keep the virus in check after primary infection, impaired immune function due to stress, allows the virus to multiply and results in increased antibodies (Glaser 2005; Kiecolt Glaser and others 2002a; Kiecolt Glaser and others 2002b;
McDade and others 2000a). A wide body of evidence supports the use of EBV antibodies as a sensitive measure of psychosocial stress, or more precisely, HPA and sympathetic activation
135 (Glaser 2005; Kiecolt Glaser and others 2002a; Kiecolt Glaser and others 2002b). In addition, the development of a blood spot assay for EBV antibodies means that samples can be collected with a minimum of time and discomfort for participants, and that samples can be easily stored in field conditions (McDade et al., 2000a). Potential confounders of EBV antibody concentrations include current infection and nutritional status (Chandra 1997; Kiecolt Glaser and Glaser 1988).
Body mass index, hemoglobin, C reactive protein, and subjective illness will be tested as potential controls for these. Antibodies to EBV will be assayed in finger prick blood spots, applied to standardized filter paper as described by McDade and colleagues (2000a).
Inflammation. C reactive protein (CRP) is released by the liver as part of the acute phase response to tissue injury, and can increase by a factor of 1000 in response to a wide range of infections (Black and others 2004; Mortensen 1994), but may be otherwise undetectable in blood or sera (Fleck 1989). Low level elevations are indicative of chronic inflammation and predict severity and risk of complication from atherosclerosis (Alber and others 2008; Haidari and others
2001; Ikonomidis and others 2008; Mullenix and others 2008; Zebrack and others 2002).
Marked elevations are the result of acute bacterial or viral infection (Mortensen 1994). Plasma
CRP rises 24 – 48 hours following infection and is thought to provide non specific protection from microbes prior to immune activation (Mortensen 1994). As such, it represents an effective and objective measure of current, non specific infection; elevations of >5.0 mg/L have been used as indication of active infection (McDade 2001; McDade 2002; McDade and others 2005).
Nutritional status, particularly levels of body fat, is a potential confounder of CRP. Measures of body composition, specifically percent body fat and BMI will be tested as potential confounders.
CRP concentrations in mg/L will be assayed from finger prick blood spots, applied to standardized filter paper, following procedures outlined by (McDade and others 2004).
136 Hemoglobin. Anemia is caused by a reduction in hemoglobin, the molecule that carries oxygen in the blood. This in turn means that less oxygen is being delivered to tissues. The most important cause of anemia world wide is lack of iron in the diet (Hunt 2002; Lehti 1989; Tatala and others 1998). The effects of anemia include increased morbidity, mortality, growth stunting, impaired cognitive ability, reduced work capacity and complications in pregnancy (Hunt 2002;
Lehti 1989; Tatala and others 1998). In addition to diet, hemoglobin concentrations are elevated among high altitude Andean populations, necessitating different cutoffs for defining anemia than at sea level (Beall 2001; Beall and others 1998). Andeans are also at greater risk of complications from high hemoglobin concentrations. Recommended cutoffs for identifying anemia among women living around 4000 m, based on slightly different methodological approaches are 14.5 g/dL (Cohen and Haas 1999) and 12.0 g/dL (Vasquez and Villena 2001).
Based on consultations with a local doctor, we used a cutoff of 13.0 g/dL. Abnormally high levels of hemoglobin, or polycythemia were considered those 19g/dL or higher (Vasquez and
Villena 2001).
Hemoglobin levels were tested through finger prick blood samples, in a HemoCue B
Hemoglobin photometer, according to the procedures outlined in Tatala and colleagues (1998), and reported in g/dL.
Data analysis
Laboratory analysis. Blood spot samples were analyzed by the student PI at the
Laboratory for Human Biology Research at Northwestern University using enzyme linked immunosorbent assay (ELISA) according to previously validated protocols.
Assay of EBV antibodies. The following protocol was developed and validated by
McDade and colleagues (2000a), to use a commercially available assay kit (No. 7590, DiaSorin
137 Corporation, Stillwater, MN), to measure p18 VCA IgG antibodies to EBV. One day prior to analysis, a 2.5 mm hole punch is used to remove a sample of the blood spots, and these are eluted in 250 L of diluent buffer, overnight at room temperature. 100 L of eluate is added to each well, and the assay completed following kit instructions. The kit includes four levels of p18 VCA IgG standard, and three levels of control stabilized human sera. Four additional levels of control in whole blood are added to filter paper, dried overnight, and stored at 23˚C.
Standard concentrations, supplied by the manufacturer, are used derive standard ELISA scores from absorbance values.
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; that is, it is designed to investigate variation primarily at the lower end of the distribution. Higher concentrations are assessed by further diluting the sample to put it in the detection range of the assay.
Statistical analysis. All analyses were performed using STATA 8.0 (Stata Corporation,
College Station, Texas).
All continuous variables were tested for normality by visual assessment of the distribution, and statistical tests (skewenes and kurtosis, and Shapiro Wilk tests for normality). Where possible, non normal distributions have been log transformed to preserve continuous variation. Where log transformation did not lead to improvement, variables were divided into categories and modeled as a dummy variable.
Multiple linear regression analyses were used to explore the relationship of household composition, social support and SES to each of the health outcomes (i.e., percent body fat, hemoglobin, CRP, EBV), controlling for appropriate confounders. BMI and CRP were log
138 transformed in order to normalize their distributions. In testing for the influence of household composition on health, a dummy variable was created for female headed households.
Significance levels were set at P < 0.05. With regard to SES, ratios of the number of economically active individuals, and rooms (room ratio RR, work ratio WR) in the house were divided by the number of people in the household, to create a ratio, and then log transformed to normalize their distributions. Neighborhood (school, health clinic, hospital) and household characteristics (water, plumbing, electricity, gas stove) were added, and then dichotomized into two groups with 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 household. Finally, access to social support has been modeled as a series of dummy variables, which will be further explained below.
139
Chapter 5: Household Structure, Social Relations and Health Status in El Alto
Introduction
In this chapter I present descriptive statistics and bivariate correlations both to investigate
variation within this sample with regard to household composition, SES and social support, and
to test associations which may be relevant for hypothesis testing; more specifically, to determine
which variables should be added as controls in multivariate models. As in the previous chapter, I
will begin with interview data, and finish with biological outcomes. In the next chapter, I will
present multivariate models developed based on these associations.
Instrumental variables
Household composition
Table 5.1. Number of total individuals and economically active individuals by women’s age. Household size Number of workers Age N Mean S.D. Min Max Mean S.D. Min Max 18 29 36 7.4 4.0 3 20 2.9 1.2 1 6 30 39 29 6.9 3.7 4 21 2.4 1.0 1 6 40 49 12 6.6 2.6 3 12 2.6 1.3 1 5 50 75 14 6.4 2.8 2 10 3.3 2.2 1 10 Total 91 7.0 3.4 2 21 2.8 1.4 1 10
Table 5.1 displays means and standard deviations for the total size and number of
workers for households of different ages of study participants. Older women have smaller
households, with more economically active members, but these associations are not significant.
Among the total sample, the average household size is 7 individuals, with 2.8 economically
active members. Women had an average of 3.3 children, but among the older age categories,
140 they are less likely to be current members of the household. As we might expect, the number of children is positively associated with age, or, perhaps more appropriately, women’s reproductive lifespan (Spearman’s Rho 0.49, p <0.000). Two (2%) of the participants were pregnant at the time of the interview, 16 (18%) were lactating, and 18 (20%) are post menopausal. Finally, those participants who return home to their rural community (N=70) go home on average
(geometric mean) 1.8 times per year (CI 1.3 2.4), generally to visit, and very often to help with the yearly harvest and bring home produce.
Table 5.2. Household headship. Household Frequency % Age (SD) DHH 50 55 35.4 (12.2) FHH 14 15 44.1 (12.2) MHH 20 22 34.7 (8.6) Living with parents 7 8 25 (3.5)
Table 5.2 displays the frequency of women with different living arrangements. Over half of the women reported that they share economic responsibility and decision making with their husband. Fifteen percent of women reported that they are exclusively responsible for household economic production, including a woman whose husband lives in Argentina, and another whose husband cares for his father in a rural village. Relatively few women reported that their husband was in charge of the household or that they were single and living alone, twenty two and eight percent respectively. Of those women living in an extended household with their parents, three had children, and four did not. There are significant differences in age between these different categories (Pearson’s chi2=12.8, p=0.005): women heads of household are, on average, older, while women living at home are younger. There was no significant association between household headship and the ratio of workers to total household members (Pearson’s chi2=1.6,
141 p=0.655); the ratio of the number of rooms to total household members (Pearson’s chi2=2.5, p=0.468); or the total number of rooms in the house (Pearson’s chi2=3.9, p=0.277).
Social support
In the preliminary interviews, we asked the participants what kinds of problems they typically need help to solve. The most common answers among the sample of twenty nine needed help, particularly financial help, with an illness, needing money, and needing advice regarding family or interpersonal conflict. We also discussed with participants the problems that cause them the most worry or preoccupation. Again, the most common problems needing money, providing 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.
Table 5.3. 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 receiving 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
Of the sample of ninety one women, most have family in the area, and three quarters have compadres and/or padrinos . Table 5.3 displays the number and percentage of women with compadres and padrinos with perceived instrumental or emotional support from these relationships. In each case, emotional support is more common than instrumental support, and support from kin more common that from fictive kin. The mean number of the 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), participants reported an average of 0.9 (SD 0.6, min 0,
142 max 2) and 1.5 (SD 0.9, min 0, max 5) respectively. Twenty one (23%) of women reported no source of instrumental support, while only 8 (9%) reported no source of emotional support.
In the initial ethnographic interviews, we asked the women to discuss their compadrazgo relationships; particularly whether they perceived any benefits associated with these connections.
Most of the women in this small sample of twenty nine reported that they either did not have compadres , or saw them only rarely, for example, once a year at Christmas. While one woman felt that having compadres was simply a formality associated with baptism, several others conveyed that they shared an affectionate and mutually supportive relationship with their compadres . 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, sixty six of the ninety one women 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 Catholic and Protestant had compadres. Most of these ritual ties were associated with baptism, but there were others, for example, compadres for a child’s graduation. Fourteen women described their relationship with their compadres as emotionally supportive, and twenty three their padrinos . Eleven women reported that they receive some kind of economic or instrumental support from their compadres , while only eight said the same of their padrinos.
143 Table 5.4. Number of women who needed and received help for specific problems; number and percentage of women needing help who received it, and; number and percentage turning to specific sources of help. Needed Received Problem help help Family Friend Neighbor Compadres Padrinos Other N/91 N (%) N (%) N (%) N (%) N (%) N (%) N (%) Find work 84 78 (92) 26 (31) 28 (33) 9 (11) 1 (1) 2 (2) 12 (14) Find lodgings 65 65 (100) 45 (69) 7 (11) 4 (6) 3 (5) 0 6 (9) Childcare 71 58 (82) 45 (63) 2 (3) 2 (3) 0 0 9 (13) Illness 74 73 (99) 31 (42) 1 (1) 1 (1) 2 (3) 0 38 (52) Loan 45 39 (85) 10 (23) 1 (3) 2 (5) 0 0 25 (57) Conflict w Husband 43 30 (68) 20 (47) 4 (9) 2 (5) 0 4 (9) 0 Conflict w M in L 17 11 (65) 5 (26) 0 0 0 0 5 (25)
Table 5.4 presents information on need and access to social support in the past. These
problems did not all apply to all women; for example, women without children did not need help
with childcare, along with a smaller number of mothers. Most people who needed help received
it, though this was slightly less true for women having trouble with their husband, money or
childcare. In all but one case (work) the majority of participants turned to family before friends,
neighbors, compadres , or padrinos . With regard to having an illness or needing a loan, a fairly
large number of women said they would only be comfortable seeking help from a doctor or from
the bank. Conflict with the mother in law, although it was mentioned several times in the
unstructured interviews, was not a common problem in this group.
Table 5.5. Current, perceived support, including sources of first and second resort. Problem No one Family Friend Neighbor Compadres Padrinos Other 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)
144 Table 5.5 presents perceived support, or who women would turn to if they needed help now, regarding a similar list of problems. A fair number of women reported that they had no one to turn to if they needed money. Again, for most problems, participants were most likely to turn to family first, followed by friends. Both Tables 5.4 and 5.5 demonstrate that padrinos are a common source of support for marital conflict, but are otherwise absent. A comparison of
Tables 5.3 and 5.5 reveals that the different questions yielded slightly different results. When asked if they could ask their compadres or padrinos for instrumental or emotional support, women were more likely to say yes, but less likely to list them in response these specific problems. Finally, women were less likely to turn to any of these sources with regard, again, to illness and needing a loan, and were also more likely to turn to colleagues or supervisors for work related problems than friends, family, neighbors or fictive kin.
Socioeconomic status
Table 5.6. Neighborhood and household resources. Neighborhood N (%) Hospital 19 (21) Health Clinic 59 (65) School 75 (82) Household N (%) Electricity 84 (92) Running Water 82 (90) Plumbing 54 (59) Gas Stove 83 (91)
Table 5.6 displays the household and neighborhood resources reported by participants.
Few women lived in a neighborhood with a hospital – slightly more with a health clinic – and just over half had plumbing in their residence.
145 Table 5.7. Geometric means and confidence intervals for education, rooms in the house, and ratio of rooms and workers to total number of household residents. Education Rooms in House Room Ratio Work Ratio Age N Mean 95% CI Mean 95% CI Mean 95% CI Mean 95% CI 18 29 36 8.5 7.3 9.8 2.8 2.4 3.2 0.41 0.33 0.52 0.40 0.35 0.46 30 39 29 5.0 4.0 6.4 2.4 2.0 2.9 0.38 0.29 0.51 0.35 0.31 0.40 40 49 12 5.9 4.2 8.2 3.2 2.3 4.4 0.51 0.35 0.76 0.41 0.33 0.50 50 75 14 2.6 1.4 4.8 4.0 3.2 4.9 0.71 0.56 0.90 0.50 0.39 0.65 Total 91 5.8 5.0 6.7 2.8 2.6 3.2 0.45 0.39 0.52 0.40 0.37 0.43
Table 5.7 displays geometric means and 95% confidence intervals for years of education,
rooms in the house, and the ratio of both rooms and workers to the total number of residents.
Older women have significantly fewer years of education than younger women (Spearman’s Rho