Sociology of Health & Illness Vol. 30 No. 4 2008 ISSN 0141–9889, pp. 599–615 doi: 10.1111/j.1467-9566.2007.01079.x TALYA Sociology SHIL © 0141-9889 Original XXX COMMUNITY 2008SALANT Foundation Articles of Health EPIDEMIOLOGIES AND for &Ltd SARAH Illness the Sociology GEHLERT OFofBREAST Health &CANCER Illness/Blackwell RISK Publishing Ltd Blackwell Oxford, UK Publishing
Collective memory, candidacy, and victimisation: community epidemiologies of breast cancer risk Talya Salant1 and Sarah Gehlert2 1 2
Committee on the History of Culture, University of Chicago School of Social Service Administration, University of Chicago
Abstract
Collectively shared ideas of community may be equally relevant for the study of health disparities as quantifying the relationship between community structures and health. Data from focus groups (N = 18) that explored understandings of breast cancer and breast cancer risk in African American neighbourhoods revealed three conceptual domains where shared ideas of community informed responses: collective memory, community candidacy, and community victimisation by external aggressors. Reading the focus group responses in terms of these domains identified perceptions of risk and of candidacy that may be overlooked by individualised or quantitative approaches to studying breast cancer risk perceptions and related behaviours. These include novel perceived risks, such as the ‘risk of knowing’, as well as community-level constructions of breast cancer candidacy. ‘Lay epidemiologies’ of breast cancer within this population might therefore be better understood as ‘community epidemiologies’, where community is central to the interpretation and operationalisation of breast cancer risk. Paying attention to such community epidemiologies of breast cancer provides theoretical insights for studying breast cancer disparities and risk perceptions as well as useful guidance for designing interventions.
Keywords: African American, community, lay epidemiologies, breast cancer risk perceptions, qualitative analysis
Introduction: rethinking ‘community’ as a ‘thinking’ community Emphasis on the relationships between health and community has increased in recent years, particularly among those studying health disparities. Although differing definitions of ‘community’ have sparked debate within the health services literature (Jewkes and Murcott 1996, Wayland and Crowder 2002) much of this research continues to define and operationalise ‘community’ in terms of its visible, structural, and therefore measurable features. The prevailing focus on charting census data and mapping local institutions within the study of health disparities probably reflects its sociological underpinnings, which emphasise the health effects of quantifiable neighbourhood characteristics (Sampson et al. 2002). Additionally, government interest in and support for health disparities research might encourage emphasising variables that can be easily translated into policy recommendations. Notwithstanding the salience of built environment for influencing health outcomes, however, few researchers in this field have considered how ‘community’ might also serve © 2008 The Authors. Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd. Published by Blackwell Publishing Ltd., 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA
600
Talya Salant and Sarah Gehlert
important interpretive functions for how individuals navigate the landscape of health. In this sense, community is an idea collectively shared by the individuals it purportedly circumscribes that is used to explain and ultimately produce particular health behaviours observed at the sociological level of community. Benedict Anderson (1991) described how ‘imagined communities’ can generate sociopolitical structures and collective action even in the absence of geographic boundaries. Therefore, a community’s ongoing existence may be as much a function of how its members imagine and define it as its more measurable features – and it may even exist despite them. In the field of health disparities, studying community in terms of collectively negotiated ideas of community could help researchers link the socioeconomic and sociopolitical conditions that shape members’ lived realities to their everyday engagements with that reality. Yet, little attention has been paid to the significance of ‘imagined communities’ within health disparities research, particularly how such ideas of community might influence health attitudes and behaviours. This article uses data from a study of breast cancer health disparities within a geographically defined African American community to argue that understandings of breast cancer risk and related health behaviours are filtered through collectively shared ideas of community. These ‘community epidemiologies’ not only challenge operational definitions of community within health disparities research – and in so doing expand the meaning of ‘risk’ – but they also provide useful guidance for designing interventions.
Methods This article uses narrative data from 18 focus groups that were conducted on-site in nine predominantly African American neighbourhoods on the south side of Chicago. These were a randomly selected subset of 49 focus groups conducted in 15 neighbourhoods as part of a larger collaborative effort to explain local breast cancer survival disparities (Figure 1). These disparities mirror US trends, where black women experience excess breast cancer mortality when compared to non-Hispanic white women, despite lower overall incidence rates (Newman 2005). As the exploratory component of this multi-tiered project, the focus groups were intended, broadly, to ‘identify breast cancer issues important to local community residents,’ while building collaborative partnerships and networks to further examine ‘the relationship between psychosocial factors and health outcomes’. Other research tiers, as shown in Figure 1, attempted to measure community structures and psychological or biological variables that might predict poorer survival. Focus groups consisted of open-ended questions related to breast cancer understandings and experiences (Figure 2). Three of the questions in the focus group battery were explicitly designed to explore ‘lay epidemiologies’ of breast cancer risk, including: meanings of ‘at risk’, causal attributions, and ideas about candidacy – i.e. who gets breast cancer and why (Davison et al. 1991, Frankel et al. 1991). Adapting a conceptual model introduced by Davison et al. (1991) and Frankel et al. (1991), the aim was to better understand the sophisticated health explanatory models underlying individual behaviours and attitudes about risk and prevention observed elsewhere in breast cancer high risk clinics (Salant et al. 2006). That is, beyond their insights into community attitudes towards and understandings of breast cancer, the focus groups were regarded as an ideal methodology to enable the emergence of shared cultural explanations and logics of breast cancer risk that may be constrained by the power-laden setting of the medical encounter or by one-on-one interviews. © 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
Community epidemiologies of breast cancer risk
601
Figure 1 Collaborative health disparities project model
Figure 2 Sample focus group questions
Recruitment occurred through advertisements in local newspapers, poster and flyer distribution by project staff within neighbourhoods (at bus stops, on streets, in parking lots), postings at health clinics and aldermen’s offices, and announcements at local gatherings (church services, meetings). Of 1,300 interested volunteers, 503 were selected to form two © 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
602
Talya Salant and Sarah Gehlert
to three groups per neighbourhood area that: (1) represented the demographics of the area, and (2) would result in groups heterogeneous in terms of age, gender and socioeconomic status, while avoiding situations in which some person(s) in the group would be dominant over others (e.g. including a participant’s work supervisor or having a mother on public aid in a group otherwise made up of professional men). As a result, focus group participants (7–13 per group) were predominantly African American adult men and women of varying age, gender, employment status, marital status, religion, and personal or family experience of breast cancer (see Appendix). During the two-hour sessions, trained moderators used the open-ended questions as a guide, frequently modifying question order and wording and providing probes to explore participants’ responses. Groups were tape-recorded and the data professionally transcribed. Since question responses were not expected to be mutually exclusive, focus group sessions were treated as cohesive wholes and the randomly selected subset of 18 transcripts coded and analysed – using NVIVO software (QSR International©) – in their entirety by one of the authors (TS) in consultation with the co-author (SG). Coding proceeded by using the question guide as a general framework to create broad coding domains (e.g. responses to ‘What women do you think are most at risk for getting breast cancer?’) and then proceeding iteratively, according to methods outlined by Strauss and Corbin (1994). While guided by a broad theoretical objective (i.e. describing lay epidemiologies of breast cancer) the latter approach identified concepts or themes that overlapped across question domains (e.g. causality, nostalgia, ‘worriation’, secrecy, etc.). As a consequence of this ‘grounded’ analytic approach, which does not define important conceptual domains – such as risk – a priori, the interpretive significance of community became increasingly apparent within lay epidemiologies of breast cancer. That is, interpretations of breast cancer risk emerged hand in hand with a socially constructed idea of community that circumscribed the visible geographic boundaries of the different neighbourhoods, suggesting further that individualised operational definitions of risk may overlook the ways that community epidemiologies of breast cancer underlie individual attitudes and behaviours. The remainder of this article, therefore, uses a community (vs. individual) lens to describe the languages of risk – the community epidemiologies – that emerged in the focus group discussions. It does so by presenting three conceptual areas arising from the data analysis – collective memory, candidacy, and victimisation – that illustrate the symbiotic relationship between community as imagined and expressed understandings of breast cancer risk. As might be expected, the distinction between the three concepts is somewhat illusory, since they demonstrate considerable overlap, particularly in how they jointly explain behavioural patterns such as low mammogram screening rates or diagnostic delay. However, in representing one particular slice through rich narrative data on community experiences of and understandings of breast cancer, they are useful for broadening the operational use of community in health disparities research, for expanding the interpretive contexts of breast cancer risk, and for offering practical lessons for intervention. They also identify active and constructive strategies of interpretation that challenge a tendency to view behaviours within particular – ethnic, non-white – communities as deviant or as the result of ignorance (Abrums 2000). The following discussion presents the concepts that emerged from the coding of focus group transcripts, using for illustration representative quotes and tabular representations of theme distribution. Data in tables is presented as categories of response to particular questions and as tabulations of dominant themes (e.g. causality). Respondents’ quotes are taken from 14 of the 18 focus groups and were modified only where necessary to clarify their meaning. © 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
Community epidemiologies of breast cancer risk
603
Collective memory Nostalgia was a common theme throughout focus group transcripts, occurring seventh in frequency out of 173 codes. In particular, participants frequently waxed nostalgic about traditional ways of life when discussing breast cancer causality – usually in the context of explaining the meaning of or types of people ‘at risk’. Some referenced old-fashioned lifestyles when identifying unhealthy dimensions of modern lifestyles, most frequently diet: I think more should be done on what people actually eat nowadays because I know when I was a young boy, it was greens and corn bread and actually eating meals. Nowadays, your meal is coming from Burger King, McDonalds, Wendy’s, that’s dinner (1703). Microwaves. That’s the reason I think there’s so much of [breast cancer] happening with young people because they like to be out and they party, they come in and they don’t want to cook anything, they just throw everything – it goes in the microwave and you know it’s full of radiation (4106). In many of these responses, the past connoted a more natural time or place (e.g. the South) where the source of and knowledge about foods was predictable and controllable. In contrast, modern foods were regarded with suspicion as unnatural and physically harmful: You know, like when I was little, we ate everything under the sun – I was born and raised on a farm. We would go straight out in the field and just pick something – and not wash it or nothing. But you know, they got so much stuff – they say, ‘don’t eat this.’ Every day you wake up and on the news, they say, don’t eat this, don’t touch that – so what do you do? (0604). See, I have a family that has a farm. My grandmother used to just go out in yard and grab at chicken and bang! That was fresh chicken. And he wasn’t big as this chicken is right now. And now that chicken was fresh. Now, it’s all type of chemicals in these chickens (0202). These passages emphasise chemical and/or environmental sources of risk in food rather than individual dietary choices per se. The findings triangulate with causal theories of ‘external’ risks to health, discussed later. The implicit link between both sets of arguments is a sense that modern times threaten natural balance and control, whether through the fast pace of living that necessitates unhealthy eating (e.g. microwaves) or indirectly through the modern industrial use of harmful chemicals to increase food production. Nostalgic references to the South may reflect the migratory history of black Midwesterners many of whom trace their family history to the ‘Great Migration’ and maintain kinship ties to small towns in Mississippi, Arkansas, and Alabama. Resemblances to biblical parables of corruption notwithstanding, these juxtapositions of modern (urban) risks and traditional (rural) health also evoke the notion of ‘civilization as risk’, that Rosenberg (1998) observed to be a trans-historical explanation about the causes of illness. Cyclical in popularity, this holistic view of the relationship between one’s environment and physical health has represented both a reaction against the reductionist tendencies of modern medicine and a commentary on social class inequalities (Rosenberg 1998). Its appearance here suggests that community epidemiologies of risk may also contain implicit political critiques, as observed elsewhere (Balshem 1993), which cannot simply be explained as © 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
604
Talya Salant and Sarah Gehlert
fatalistic beliefs. No doubt, such theories reflect lay associations between unhealthy living environments and poor health, links that are infrequently examined in breast cancer epidemiology (Krieger 2005). Apart from their direct causal explanatory value, references to traditional or old-fashioned ways of life were also used to explain other kinds of risks related to breast cancer that emerged out of the lack of widespread access to and utilisation of available medical care. Here, questions posed about risk and causality as well as questions that accessed personal concerns about cancer (e.g. ‘How does breast cancer fit into your day-to-day life?’ and ‘What comes to mind when you think about breast cancer?’) elicited responses that indicated a complex and frequently contradictory set of explanations about why individuals do not obtain medical attention when they should. In particular, participants attributed widespread denial and ignorance to the fear and social stigma of cancer perpetuated across multiple generations. It’s the perception and I think it could span . . . across generations. When you hear that word: cancer. You know, 50 years ago if you heard the word cancer, it was an automatic death sentence (4108). So much has been hidden . . . [Cancer] was a word that you just didn’t even use. And you said, ‘well they got the big “C” ’. And then all of a sudden that person became isolated. It was like it was contagious to be around that person. . . . And so you grow up with some of these myths in your head and as soon as you hear someone with cancer you expect the worse, you know (2215). But that fear is going to keep us sick. The fear of not talking to each other . . . [S]ome women aren’t comfortable with themselves enough to do their self-checks. And I can’t think of one brother I know my age, that would have a conversation about prostate cancer. So we got to get this fear – the African American community – we got to get on this fear because fear is killing us all (3503). Here, lay epidemiologies of breast cancer morbidity and mortality, or the confounded risks of acquiring breast cancer and of dying from the disease, characterised participants’ responses to a variety of questions about breast cancer. That is, thinking about breast cancer relied on a broader social and cultural analysis than simply explaining incidence. Participants acknowledged that the common experience of family members and friends passing away from breast cancer, particularly among the older generations, reinforced the sense for many that breast cancer was invariably fatal and therefore frightening to think or speak about. . . . [S]ome of them, if they don’t lived through – up to a certain age – they done witness a lot of sickness in their family – some people just feel like this, they would rather not know, ‘I’m healthy now, I feel healthy, what I don’t know won’t hurt me’. They think like that, a lot of them do (0908). That’s because other people done told her, ‘well you know, there’s a risk to [the treatment].’ That’s why some people keep it to themselves. They are afraid of [it]. . . . and its nothing but hearsay, and they don’t go and check in on it for themselves. Because my grandfather died of cancer too – we knew – but we couldn’t make him (2805). [T]hey think if they don’t talk about it, it may go away (1700). © 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
Community epidemiologies of breast cancer risk
605
In short, beyond explicit ideas about the causes of cancer, understandings of breast cancer risk also incorporated the perceived ‘risk of knowing’ that participants recognised as a frequent barrier to seeking medical attention, whether for cancer screening or to address an ongoing physical problem. The ‘risk of knowing’ and its associated tendencies towards silence and intentional ignorance (i.e. choosing not to know) were seen as created and perpetuated by an older generation that experienced or witnessed overwhelmingly negative outcomes from receiving breast cancer diagnoses. Even though some pointed to improvements in breast cancer treatments, participants noted that the fear and silence among this generation fed the cancer attitudes and unawareness of younger generations. As a novel risk category, the risk of knowing also triangulated with themes emphasising emotional and cognitive causes and consequences of breast cancer. Whether as stress causing cancer, fear that dwelling on cancer might cause it, or belief in the power of the mind to heal the body, ‘worriation’ (Snow 1993) was the 13th most commonly mentioned theme.1 Perceived emotional causes of cancer and the reluctance to ‘dwell on’ one’s risk similarly characterised African American participants at breast cancer high-risk clinics (Salant et al. 2006). As discussed later, worriation – manifested through the risk of knowing – may also help explain the frequently expressed lack of daily concern with breast cancer. Additionally, the risk of knowing was viewed as a product of negative collective experiences with a mistrusted medical system, seen itself as a source of risk to one’s health. Complaints of inadequate or unequal care – the most commonly coded theme – proliferated in responses to questions about access to treatment and screening. This frustration was based in widely recognised contemporary healthcare inequalities, where the economically disadvantaged and uninsured receive suboptimal care and confront numerous logistical barriers in obtaining even the most basic of medical services. As some pointed out, however, backgrounding these complaints was the persistent memory of historical injustices against African Americans (e.g. the Tuskegee syphilis experiments) incurred by a prejudiced and patriarchal medical system: I remember. . . . my uncles were born in 1895 and 1910 – and I think about how black folks – from that time – had a cultural aversion to doctors because most of the doctors were white. And really were insensitive to the needs of black folks. They didn’t have to be racist but they could be culturalised not to be sensitive to us. And then you turn around in the twenties and thirties and you got the Tuskegee experiments – and I had relatives and to them that was current events – so I think a lot of this was the shame of being sick with something that they can’t deal with. A lot of people feel shame. Not that they did something wrong, but to be able to handle it or can’t think of a way or lose the way to discuss it with someone who might want to know (2201). I think you are speaking to a real important point in the African American community, at least – is our distrust for the medical system. For our doctors and physicians and nurses – we don’t trust them . . . And I think because we don’t trust them, we don’t go to them and despite information – when we talk about education – despite what we hear – and you even said, we should have just one message – and I think that would really simplify things, but I think at the same time, what works for one community, may not work for our community because of that distrust. So I think, that’s some real important component in terms of dealing with something that we may have. I think that once we get a diagnosis or we are at risk for it, that we’re not as involved actively in dealing with it as we could because of our distrust (2006). © 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
606
Talya Salant and Sarah Gehlert
In sum, collective memory informed the ways focus group participants made sense of breast cancer risk in several ways: as a referent for identifying the uncontrollable and unnatural dimensions of modern (vs. traditional) ways of life, as the cross-generational lay epidemiologies of death and fear that reinforced perceived risks of knowing one’s health status, and as the legacy of targeted harmful medical treatment (i.e. medicine as risk) that continue to hinder individuals from seeking medical care. Sometimes nostalgic and sometimes accusatory, references to a collective past nonetheless provided a constant resource with which the respondents organised and re-produced the meaning of breast cancer risk. The widespread use of collective memory suggests that as long as the social and historical conditions sustaining collective memories of nostalgia, injustice, and death endure, breast cancer risk will continue to be understood in terms of modern dangers, medical mistrust, and the risk of knowing. In particular, as long as mortality from breast cancer remains disproportionately high in certain populations, feeding community epidemiologies of breast cancer risk, these collective memories will continue to motivate individual attitudes and health behaviours. Far from urban legends or markers of ignorance, therefore, the use of collective memory reflects sophisticated responses to epidemiological and social realities while ironically serving to reinforce these realities by encouraging intentional ignorance and related health behaviours (e.g. diagnostic delay). Lastly, the usefulness of collective memory as an organising conceptual device may also help to explain an apparent paradox in the focus group discussions. That is, respondents continually insisted on the need for greater medical information and education about breast cancer even as they acknowledged the unwillingness of community members to embrace or act on available information. This observed contradictory desire for and fear of knowledge can be understood by tracing the different ways collective memory was invoked in responses to questions about breast cancer risk. When it referred to the ongoing legacy of injustices, collective memory identified the persistent lack of equitable, unbiased medical care; when it referred to the cross-generational lay epidemiologies of death and fear, however, it underscored the reluctance of many openly to embrace information about their health (the risk of knowing). That both sets of explanations could be and were invoked demonstrates the flexibility and resilience of the concept. It also suggests that individual behaviours (e.g. non-compliance or diagnostic delay) or risk perceptions (e.g. ‘I don’t dwell on it’) observed in clinical contexts (Salant et al. 2006) might be rooted in a more complex set of community experiences and meanings overlooked by individualised clinical or research strategies that focus on quantified perceptions of risk. Further, it identifies possible responses to new forms of health risk information – e.g. genetic risk – that may be regarded as simultaneously desirable and harmful. Those interested in making sense of and deriving practical interventions from community understandings of breast cancer must therefore first understand how shared concepts such as collective memory structure those understandings, particularly of risk, and create an implicit logic beneath apparently contradictory attitudes.
Candidacy Like collective memory, the concept of candidacy emerged in the ways focus group participants made sense of questions about breast cancer risk. These included questions assessing participants’ ideas about the types of women most at risk for breast cancer as well as those exploring competing daily risks and the ways that particular communities are constructed as candidates for different diseases. While the first type of question addresses © 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
Community epidemiologies of breast cancer risk
607
general ideas about which individuals might be at risk for breast cancer, the latter two address the community and everyday contexts in which breast cancer risk acquires – or loses – visibility. This community-level understanding of candidacy has arguably received little attention within risk perception and health disparities research. Davison et al. (1991) previously introduced the concept of ‘candidacy’ within lay (i.e. non-medical) epidemiologies of heart disease, where it was used to uncover lay explanatory models about causality, specifically about ‘the kind of person who gets heart trouble’ (p. 5). Ideas of candidacy, they explained, are ways of particularising general knowledge of disease, or of drawing community experiences and understandings of illness into an individual frame. In their analysis, notions of candidacy referred primarily to causal agents applicable to individual-level characteristics or behaviours (e.g. fat people, worriers, smokers, etc.). Similarly, focus group questions about the types of women most at risk for breast cancer elicited responses about individual causality – for example, heredity, lifestyle (smoking, drugs, alcohol, diet), age – as well as responses that revealed the interrelation of breast cancer morbidity and mortality within understandings of risk – for instance, not seeing a doctor, poverty, unequal care (Table 1). A fair number claimed that ethnic identity influenced likelihood of cancer; however, ascriptions of risk to different ethnic groups (Whites, Asians, Blacks) ranged widely while others countered that race was not a factor – all women are at risk for cancer. When invoked, Black ethnicity was often a proxy for dietary preferences or a lack of social opportunities that made its members more vulnerable to poor diets or environmental risks; it was not a commentary about inherent racial differences per se.
Table 1 Focus group response categories and frequencies Response categories
Times mentioned
A. ‘What women do you think are most at risk for getting breast cancer?’ Don’t see regular doctor/get regular checks Ethnicity (e.g. Black, White, Asian, etc.) Low income/disadvantaged/poor Age (older, younger) Environment/chemical exposure Diet Drugs, alcohol, smoking All are at risk Unequal care Low education Taking hormones/medications Hereditary Worry, stress Large breasts Bruise/trauma to breasts
31 30 28 24 21 19 15 14 14 14 13 9 8 8 6
B. ‘When you think about things that concern you on a day-to-day basis, where does breast cancer fit in?’ Don’t think about it /bottom of list Other life risks a priority Lack of public info on breast cancer Think about it daily Depends on personal experience with cancer
32 23 9 11 23
© 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
608
Talya Salant and Sarah Gehlert
While identifying as at risk those who do not attend regular check-ups engaged a broader sociological explanation for individual behaviours, including the lack of resources and the aforementioned risk of knowing, doing so also led to discussions about the deterring influence of competing life risks. That is, while respondents seemed to feel that, on an immediate level, individual or socio-economic factors played a role in developing – or dying from – breast cancer, other constructions of candidacy came into view when respondents were asked about their personal daily concern with breast cancer and about the kinds of diseases they feel receive most attention in their communities (Table 1). That is, breast cancer appeared not to be a priority for most of those responding to questions about daily concern with breast cancer. The most common reason offered for this lack of concern highlighted that, for many members of the community, individual concern about breast cancer pales in comparison to the competing risks of daily survival, particularly in light of multiple responsibilities related to caring for and supporting their families: (Emphasis added): And where your kids are. Can they make it home too? That’s all. You’re not worried about breast cancer. And it’s something that should be addressed but it’s not . . . [Y]ou’re talking now, racial but also economics. Lincoln Park, Bellwood, Lincolnshire, they can talk about breast cancer, okay (3812). (Emphasis added): . . . [W]hat tends to happen is, we struggle economically, we have education challenges, job challenges and we have all these factors – violence, substance abuse and HIV – and even with HIV, people don’t want to deal with that because they deal with day to day life and so what happens is, I think, in our community, [breast cancer] is not talked about and its not dealt with. In our own homes, it may be because of someone who we were close to us. But, for the most part, it’s not (2006). (Emphasis added): I think on a day-to-day struggle in the minority community you are thinking about providing food for your children – whether or not you are going to have shelter, with the economy being the way it is now, cancer sits on the back burner. It only comes up when it becomes personal (2507). Only the elitists have time to think about themselves. Minorities – those struggling have to think about everything (2500). And they become desensitised, okay, another disease – something else killing us (2910). In other words, unless it hits home and then becomes part of the lived reality of daily struggles, dangers, and worries (and sometimes not even then) the possibility – risk – of breast cancer is not a recognised category of thought. By extension, the competing risks of daily living appeared to preclude personal identification with breast cancer through the risk of knowing, where respondents noted an emotional reluctance to openly identify oneself as a potential candidate for breast cancer: I think with black people, it’s more of an emotional thing. They don’t want to discuss those feelings . . . That’s like going to a funeral and then discussing it some days later – bringing back feelings that you didn’t want to feel – I think it’s more of that, than anything else. They don’t want to sit there and talk about something that’s so hurtful to them (1707). © 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
Community epidemiologies of breast cancer risk
609
I think it’s possible that they don’t share it because like within my family, you know, there’s cancer and they don’t want to talk about it. You know it, but you don’t want to face it. Whatever happens to you, you deal with it. [I]t’s very emotional (4102). Thus, while respondents identified the types of individuals at risk – i.e. candidates – for breast cancer in terms of individual and social determinants, when the discussion focused on everyday lived experiences and self-identifications with the likelihood of breast cancer, competing life risks and emotional coping strategies effaced personal candidacy. Some respondents, as cited above, insisted that imagining oneself as a candidate for breast cancer was a luxury affordable only to ‘elitists’ and members of advantaged communities. These findings demonstrate that assessing individual identification with breast cancer risk – individual candidacy – must consider the relative visibility or invisibility of such personalised risk awareness within a community and familial context of everyday dangers and anxieties about survival. The relative invisibility of breast cancer was particularly apparent within conversations about available public health information. In response to questions about diseases publicised in their communities, participants recognised the absence of public information on breast cancer as compared to information on other health conditions (AIDS, STDs, hypertension, diabetes, etc.) that they encountered on a daily basis: The only advertisements I see in my neighbourhood are for HIV/AIDS treatment and I’ve also seen in magazines for herpes and HIV/AIDS treatments. But more than breast cancer. It’s like a ten to one ratio (0613). In our community . . . one thing or two things that are stressed in our community are high blood pressure and diabetes. Breast cancer is beginning to pick up a little bit, but it doesn’t get as much attention as high blood pressure and diabetes (0906). In my community, you don’t hear about breast cancer (4701). I think of Suzanne Somers [who had breast cancer], I’ve seen several of these people weren’t, they were movie stars and that’s still another attachment. I’m not seeing Susie from the Hood. I’m struggling with breast cancer, I don’t have the money to pay for – I’ve never seen any commercials for it (2500). Greater public visibility of particular diseases in the African American community in part reflects epidemiological trends, where the incidence of diabetes, hypertension, strokes, and HIV exceeds that among non-Hispanic Whites (CDC 2005). However, some have also argued that the personal salience of particular diseases or risk perceptions can be linked to a public context of strategic illumination and concealment that serves normalising political interests, however unintentional (Eckman 1998, Lupton 1994, Press et al. 2000, Treichler et al. 1998). Although it is impossible to know whether such interests are at work here, some participants nevertheless associated the absence of breast cancer publicity in their communities with their economic and political disempowerment: If you are rich and you have the money then; your disease is prominent because you are the powerful person in the neighbourhood or in the community. So you are in the news at that moment. You can have a hang nail; but that’s big news and they done found out that you can get gangrene from having a hangnail or something because a rich person had it (4705). © 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
610
Talya Salant and Sarah Gehlert
Somehow I think, they think that breast cancer is an elitist disease and those other diseases are towards poverty because we can’t eat the way – we can’t afford to eat. I can’t eat vegetarian if I’m getting $141 a month on food stamps (2500). To the extent that powerful outsiders’ constructions of community candidacy are internalised and enacted by the members of those communities, assessing individual understandings of breast cancer risk and related behaviours must incorporate community-level perceptions of salient health concerns. In particular, the noted public absence of ‘Susie from the Hood’ beckons researchers studying breast cancer risk perceptions to examine how such silences and invisibilities structure their data. As others have shown, survey or interview responses such as ‘I don’t know’ or ‘I don’t think about it’ or failures to respond altogether may be their own sources of ‘data’ about underlying lay or community risk epidemiologies that standard labels of denial or ignorance may overlook (Balshem 1993, Rapp 1999). More insidiously, this points to self-reproducing cycles of public and private invisibilities within breast cancer research. That is, as long as African Americans are not regarded – and do not regard themselves – as candidates for breast cancer, clinical research models and interventions will continue to reflect the white, affluent individuals who populate the clinical spaces where research data are generated.
Victimisation The discussion of ‘collective memory’ identified a historically grounded sense of injustice that served as a consistent and ongoing resource for explaining observed patterns of disease and health behaviour in the African American community. It also pointed to areas where ideas about breast cancer risks were bound up in notions of ‘external’ causality. These conceptual resources recurred within causal explanations for breast cancer that appeared connected to a shared sense of community disempowerment and feared external aggression. Ideas about causality punctuated the focus group transcripts and did not always directly refer specifically to breast cancer. For instance, causal theories appeared in conversations on the topic of puberty and what participants overwhelmingly perceived as unnatural and premature physical development among young girls today. In these discussions, which centered on a theme of natural processes out of order, participants implicated various types of chemicals (hormones, steroids, preservatives) present in modern foods, particularly meats and milk products. Implied here is an underlying explanatory impulse to explicate and restore imbalances regarded as simultaneously physical and societal. In some cases, individuals associated patterns of breast cancer incidence directly with such unnatural physical development: (Emphasis added): When I first heard of breast cancer, it was something that 60-year-old women got, okay. Then it was 45 you will have breast cancer. Now here we are in the nineties and two thousands and we’re talking about people who are 15, 16, 17, 18, 19 and 20 with breast cancer, metastasised. We are talking radical mastectomy, double radical. And these girls aren’t even in college yet. That tells me something – when you mentioned environment and all the chemicals that are out here – I’m thinking part of it is some kind of chemical. And then you are talking about also, all the hormones that they are putting in the meats over the years – that we have been eating – and the kids are infested with. . . . eating meat and drinking milk from hormone cows and stuff and hormones are like the most powerful chemical in your body – that’s what switches all the cells to start growing and © 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
Community epidemiologies of breast cancer risk
611
setting down. I’m thinking, what if that’s what’s throwing it out of whack. Here she is, 14-year-old girl looking like she’s 30 (2201). Reminiscent of nostalgic references to an idealised past to explain contemporary disease patterns, this passage emphasises the perceived role of harmful chemicals in one’s food. Indeed, chemicals featured as the most commonly mentioned causal category for breast cancer alone (Table 2) where ideas about chemical and environmental causes referred also to toxic waste, dangerous living conditions, radiation exposure, and even commonly prescribed medications. Some respondents introduced these environmental causal categories in the context of conspiracy theories about governmental agendas and/or disregard for their communities’ wellbeing: [T]he environmental, the pollution – if you noticed the trucks are allowed to go down our areas more so than the White communities. They got routes that they got to stick to. So I think it is (0906). There are certain things that they dump off in certain neighbourhoods (3111). These passages reiterate the sense of mistrust and fear of being taken advantage of by powerful outsiders and a general feeling that dominant economic and political powers neglect the interests of poor communities. Other markers of mistrust included accusations that scientists or the government withhold cancer cures and scientific breakthroughs, or
Table 2 Commonly stated causes of breast cancer among focus group participantsa Times mentioned ‘External’ causality: Chemicals (toxic waste, preservatives, hormones, medications) Food – chemicals/preservatives Hormones (HRT, OCT) Environment (radiation, pollution, fumes) Heredity Stress Unnatural/out of balance Don’t know Bruise, trauma to breast Trigger Total ‘Internal’/individual causality: Food – diet Smoking, alcohol, drugs Other lifestyle habits (e.g. no self-checking) Reproductive risks (breastfeeding, age at menarche, etc.) Total
58 13 13 43 41 27 19 15 9 5 217 23 27 11 4 65
a Categories were derived from specific responses to questions about the meaning of ‘at risk’ as well as from discussions about causality that recurred throughout focus group sessions. Using a typology demonstrated by Balshem (1993), causes were separated into those implying controllable (i.e. individual/internal) risks and uncontrollable (i.e. external) risks.
© 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
612
Talya Salant and Sarah Gehlert
more sinisterly, experiment directly on communities (as ‘guinea pigs’) who lack the economic or political strength to oppose them. That the number of categories and mentions of external (i.e. uncontrollable) causality shown in Table 2 exceeded those of internal or individual (i.e. controllable) causality suggests their explanatory power as they mapped onto existing and historical experiences of disempowerment. For instance, general skepticism of available healthcare and the collective memory of negative experiences with the medical system reinforced a widely held reluctance or failure to seek medical care. Thus, although some respondents insisted on the need for personal accountability, self-education, and self-care to avoid cancer, they were countered by those who reminded them of the innumerable economic and logistical barriers to achieving those objectives. In other words, community victimisation by uncontrollable outside forces dominated participants’ ideas about breast cancer causality and underlay their efforts to justify and explain existing health behaviours. These observations caution researchers or policy makers interested in designing health-related interventions to consider first how their gestures intersect with community understandings of power and authority. Moreover, a view of cancer as an entity beyond personal control may not necessarily reflect individual fatalism (Powe 1996) or indifference to prevention but rather the active synergism of historical and daily experiences on a community level – the ‘sociological imaginations’ (Farmer 1992) – that sustain and perpetuate those understandings (Balshem 1991). Deference to spiritual sources of healing and external loci of control, observed elsewhere among Black Americans (Holt et al. 2003) makes logical sense within such externalised views of causality. Realistic prevention strategies must address these community-level constructions of causality and risk if they are to be effective and, at the very least, avoid reproducing the perceptions that will ensure their failure.
Conclusions Health disparities research promises to identify and redress the causes of unequal health outcomes, particularly among minority populations, through trans-disciplinary approaches that include community participatory research (Baquet et al. 2002). This approach rationalised the collaborative study of breast cancer survival disparities that included the neighbourhood-based focus groups discussed in this article. However, if health disparities research defines and operationalises community largely in terms of quantifiable parameters, this may undercut its ultimate effectiveness. This paper argues that expanding operational definitions of community to include ‘imagined communities’ can help guide interventions and refine theoretical and methodological approaches to studying breast cancer risk perceptions. Focus groups exploring African Americans’ understandings of breast cancer and breast cancer risk revealed community to play a central conceptual role in participants’ responses. As a powerful organising construct that helped them decipher lay epidemiologies of disease (incidence and mortality) and explain health behaviours, community appeared in the form of shared memories and experiences that were seen to structure breast cancer risk and related behaviours. Here, the risk of knowing appeared as an important influence on breast cancer attitudes and behaviours, but one that may be elided by more literal attempts to measure individual risk perceptions. Through shared experiences of disadvantage and perceptions of competing disease risks, community-level understandings of breast cancer risk helped to explain the absence or invisibility of the breast cancer ‘candidate’ from © 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
Community epidemiologies of breast cancer risk
613
everyday risk perceptions. Lastly, collective experiences of disempowerment centred etiological explanations within a critique of perceived external aggressors, rather than on individual health behaviours. Although it is possible that these findings are an artifact of the data collection setting, it is more likely that the focus groups facilitated the surfacing of these extant themes, which might be otherwise obscured by the individualised and power-laden setting of the medical encounter or interview. These findings have several theoretical implications. First, they suggest that quantifying the impact of community-level variables on health outcomes may overlook the complicated ways that community meanings shape health behaviours and responses to public health information and interventions. Secondly, by emphasising external risks and community victimisation within ideas about causality, they challenge the predominant emphasis on governmentality and individual risk avoidance within cultural theories of health risk (Lupton 1993, Peterson 1997, Robertson 2001) which reflect data collected from white, middle-class populations. Thirdly, these ‘community epidemiologies’ introduce new risk categories, such as the risk of knowing, that not only provide alternatives to cognitive explanations for particular health behaviours such as diagnostic delay, but also suggest that the meaning of risk itself is contested and culturally dependent. Fourthly, prior articulations of ‘lay epidemiologies’ (Davison et al. 1991) have not accounted for the ways that candidacy incorporates both individual-level and community-level features, such as the public invisibility of breast cancer and the competing risks of everyday life. Researchers may therefore wish to expand their operational definitions of community – and of risk – and to seek lay epidemiologies of breast cancer not only in the ways that local communities negotiate the significance of official risk discourses, but also in the ways they incorporate community itself as a conceptual resource within such negotiations. Whether the community constructions of breast cancer risk presented here constitute strategic statements of resistance against established medical knowledge and authority, as has been discussed elsewhere (Balshem 1993), is difficult to know. Some may in fact argue that the self-damaging cyclical reproduction of poor health outcomes and the risk of knowing undercut the alternative perspectives on causality and risk presented by the focus group participants. Nevertheless, understanding how expressed ideas about breast cancer risk draw from and simultaneously enact a particular idea of community may still inform the practical design of strategies and research agendas to improve health outcomes. The synergism of lay epidemiologies of breast cancer morbidity and mortality and of collective memories of disempowerment, for instance, suggest the low yield of encouraging individualised approaches to preventive healthcare seeking, as well as likely resistance to treatment or research interventions perceived to come from the ‘outside’. It also hints at the ellipses of data produced by risk perception research that seeks to quantify or explain individual statements of probability without considering the other kinds of perceived risk involved in thinking about breast cancer. Those interested in reducing disparities by increasing prevention and early detection activities may wish to focus their efforts on increasing the visibility of breast cancer relative to other health risks through targeted public health campaigns – although, as some have noted, this may have the undesired side-effect of increasing anxiety (Press et al. 2000). Certainly, this article’s findings suggest that prevention is as much about the role public health and medical institutions play in reinforcing particular health behaviours as it is about active intervention. Perhaps more salient, however, for those seeking realistic approaches to minimising health disparities is the question of whether individual deliberations about breast cancer risk and prevention may be at odds with a socioeconomic and sociopolitical environment that prioritises issues of daily survival. © 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
614
Talya Salant and Sarah Gehlert
Address for correspondence: Talya Salant, Committee on the History of Culture, University of Chicago, 1213 E 53rd St., Chicago, Illinois 60615, USA e-mail:
[email protected]
Acknowledgements The authors wish to acknowledge the support received from the US National Institute of Environmental Health Sciences (5 P50 ES012382) and the National Cancer Institute.
Note 1
Loudell Snow (1993) observed how ‘worriation’ connects the problems of daily life to physical illness within African American folk health beliefs. Daily maintenance activities of prayer and spiritual coping can thus maintain mental and physical homeostasis: ‘The mind must be kept active and engaged but not become preoccupied in “worriation” or your nerves too will “go down on you” and you will fall ill’, (1993: 91). Here, action is the precondition to health, but it is a different kind of pro-activity from that implied by the mainstream tenets of preventive medicine.
References Abrums, M. (2000) ‘Jesus will fix it after awhile’: meanings and health, Social Science and Medicine, 50, 1, 89 –105. Anderson, B.R.O.G. (1991) Imagined Communities: Reflections on the Origin and Spread of Nationalism. London: Verso. Balshem, M. (1991) Cancer, control, and causality – talking about cancer in a working-class community, American Ethnologist, 18, 1, 152 – 72. Balshem, M. (1993) Cancer in the Community: Class and Medical Authority. Washington: Smithsonian Institute Press. Baquet, C.R., Hammond, C., Commiskey, P., Brooks, S. and Mullins, C.D. (2002) Health disparities research – a model for conducting research on cancer disparities: characterization and reduction, Journal of the Association for Academic Minority Physicians, 13, 2, 33–40. CDC (2005) Health disparities experienced by black or African Americans – United States, Morbidity and Mortality Weekly Report, 54, 1, 1– 3. Davison, C., Smith, G.D. and Frankel, S. (1991) Lay epidemiology and the prevention paradox – the implications of coronary candidacy for health-education, Sociology of Health and Illness, 13, 1, 1–19. Eckman, A. (1998) Beyond the Yentl Syndrome: making women visible in post-1990 women’s health discourse. In Treichler, P.A., Cartwright, L. and Penley, C. (eds) The Visible Woman: Imaging Technologies, Gender, and Science. New York: NYU Press. Farmer, P. (1992) AIDS and Accusation: Haiti and the Geography of Blame. Berkeley: University of California Press. Frankel, S., Davison, C. and Smith, G.D. (1991) Lay epidemiology and the rationality of responses to health education, British Journal of General Practice, 41, 351, 428–30. Holt, C.L., Clark, E.M., Kreuter, M.W. and Rubio, D.M. (2003) Spiritual health locus of control and breast cancer beliefs among urban African American women, Health Psychology, 22, 3, 294–9. Jewkes, R. and Murcott, A. (1996) Meanings of community, Social Science and Medicine, 43, 4, 555–63. Krieger, N. (2005) Defining and investigating social disparities in cancer: critical issues, Cancer Causes Control, 16, 1, 5 –14. Lupton, D. (1993) Risk as moral danger – the social and political functions of risk discourse in public health, International Journal of Health Services, 23, 3, 425–35. © 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
Community epidemiologies of breast cancer risk
615
Lupton, D. (1994) Femininity, responsibility, and the technological imperative – discourses on breast cancer in the Australian press, International Journal of Health Services, 24, 1, 73–89. Newman, L.A. (2005) Breast cancer in African-American women, Oncologist, 10, 1, 1–14. Peterson, A. (1997) Risk, governance and the new public health. In Peterson, A. and Bunton, R. (eds) Foucault, Health and Medicine. London: Routledge. Powe, B.D. (1996) Cancer fatalism among African-Americans: a review of the literature, Nursing Outlook, 44, 1, 18 – 21. Press, N., Fishman, J.R. and Koenig, B.A. (2000) Collective fear, individualized risk: the social and cultural context of genetic testing for breast cancer, Nursing Ethics, 7, 3, 237–49. Rapp, R. (1999) Testing Women, Testing the Fetus: the Social Impact of Amniocentesis in America. London: Routledge. Robertson, A. (2001) Biotechnology, political rationality and discourses on health risk, Health, 5, 3, 293 – 309. Rosenberg, C.E. (1998) Pathologies of progress: the idea of civilization as risk, Bulletin of the History of Medicine, 72, 4, 714 – 30. Salant, T., Ganschow, P.S., Olopade, O.I. and Lauderdale, D.S. (2006) ‘Why take it if you don’t have anything?’ Breast cancer risk perceptions and prevention choices at a public hospital, Journal of General Internal Medicine, 21, 7, 779 – 85. Sampson, R.J., Morenoff, J.D. and Gannon-Rowley, T. (2002) Assessing ‘neighborhood effects’: social processes and new directions in research. Annual Review of Sociology, 28, 1, 443–78. Snow, L.F. (1993) Walkin’ Over Medicine. Boulder: Westview Press. Strauss, A. and Corbin, J. (1994) Grounded theory methodology – an overview. In Denzin, N.K. and Lincoln, Y.S. (eds) Handbook of Qualitative Research. Thousand Oaks: Sage. Treichler, P.A., Cartwright, L. and Penley, C. (1998) Paradoxes of visibility. In Treichler, P.A., Cartwright, L. and Penley, C. (eds) The Visible Woman: Imaging Technologies, Gender, and Science. New York: NYU Press. Wayland, C. and Crowder, J. (2002) Disparate views of community in primary health care: understanding how perceptions influence success, Medical Anthropology Quarterly, 16, 2, 230–47.
Appendix: Average demographics of focus groups (N = 206) Average age (range)
Female African American Employed Married/Committed relationship Experience with breast cancer Personal experience Family member or close friend Religion Christian Other (Muslim, Jewish)
43 (18–77) % 74 100 38 30 7 61 91 9
© 2008 The Authors Journal compilation © 2008 Foundation for the Sociology of Health & Illness/Blackwell Publishing Ltd
