A genetic future for coronary heart disease?
Address for correspondence: Kate Weiner, Institute for Science and Society, Nottingham University, Law and Social Sciences Building, University Park, Nottingham NG7 2RD
This paper is concerned with changing conceptions of genetic disease. It is based on an analysis of biomedical literature and focuses on the treatment of coronary heart disease (CHD) in four published commentary papers. The aim of this analysis is to explore the ways in which CHD is constructed as genetic and the place of genetic discourses in the wider set of ideas that circulate about the disease. This analysis is then used to consider some of the claims of the geneticisation thesis (Lippman 1991, 1992). The analysis suggests that a genetic vision for understanding and managing CHD has emerged, which has many of the hallmarks of the geneticisation imagined by Lippman. However, a number of alternative and competing models of CHD are also supported within the biomedical discourse. These are related to the different disciplines with a stake in the field of CHD, and their struggles for authority. In conclusion, it is suggested that the geneticisation thesis, as a universal claim, is at odds with the diffuse and distributed nature of biomedical knowledge and practice. Rather than analysing geneticisation in a literal way, it may be more fruitful to see the thesis, itself, as a form of boundary work (Gieryn 1983).
This paper is concerned with the claim that biomedical discourses are increasingly drawing on genetic models and technologies for understanding and managing common conditions. Taking coronary heart disease (CHD) as an important example, the paper presents an analysis of biomedical commentary papers1 about the causes of the disease. This explores first, the ways in which CHD can be constructed as genetic through particular discursive practices (drawing on Hedgecoe 2001, 2002). The paper goes on to consider how the role of genetics is constructed in alternative models of CHD. The aim of the analysis is to indicate the range of ideas currently circulating about the causes of CHD and the place of genetics in this field. The findings will then be used to interrogate some of the specific claims of the geneticisation thesis (Lippman 1991, 1992) in the context of this important common adult-onset condition.
The analysis suggests that a fully ‘geneticised’ model of CHD, as imagined by Lippman (1991, 1992), has emerged. This is, however, only one of a number of competing accounts of CHD, each of which places a different emphasis on the cause of the disease and how it should be managed. In this sense, the condition can be seen as a boundary object (Star and Griesemer 1989) that is shared across a number of domains of biomedicine, but has local meanings that are specific to particular professional discourses and practices. This makes it difficult to talk of a dominant model of CHD, as this will depend on the exact context and setting. The commentary papers used in our analysis, which are drawn from three different disciplines, can therefore be analysed in terms of the boundary work they do (Gieryn 1983). The different claims they make about the aetiology and management of CHD are integral to disciplinary struggles over authority in this field. In conclusion, it is argued that the geneticisation thesis, as a universal claim, underestimates the heterogeneity of biomedical knowledge and practices. Rather than understanding it in a literal way, it may be more fruitful to see the geneticisation thesis, itself, as a form of boundary work, which feeds into a wider public health project concerned with foregrounding systemic, rather than individualistic, interventions in the field of health.
The construction of genetic disease
It can be argued that the development of molecular genetics has been accompanied by a re-construction of what constitutes genetic disease (Martin 1999). During the 1990s, a vision of a genetic future for understanding, managing and treating many common diseases emerged. This sees common conditions as polygenic and multifactorial, in which susceptibility is conferred by the interaction of numerous complex disease genes. Flowing from this, genetic research will lead to the creation of new classification systems which will aid diagnosis, and facilitate the development of predictive testing, new treatments and targeted treatments (see for example Bell 1998, Cm 5791 – II 2003, Department of Health 1995, Mathew 2001).
These developments in genetic science and the emergence of such visions have prompted a great deal of discussion by scholars in the social sciences and humanities about their likely implications. The geneticisation thesis is one of the most prominent and enduring ideas to have emerged. It was introduced by Abby Lippman in two key papers (Lippman 1991, 1992) and these provide a wide-ranging critique of the role of genetics in healthcare, discussing both prenatal screening and a discourse of ‘prediction and prevention’ for common adult-onset conditions. There are a number of specific claims incorporated in this thesis and, paraphrasing Lippman, these include:
- 1Genetics is the dominant discourse, in both professional and popular discussions.
- 2Genetic discourses are reductionist and deterministic
- 3Society is becoming stratified along genetic lines
- 4Geneticisation conditions how health problems are defined, viewed and managed: it affects concepts of normality and abnormality, privatises and individualises health risks and responsibility, focuses attention on biological rather than social conditions associated with health and leads to the increased use of genetic technologies in healthcare.
- 5Genetic discourses suggest that genetic research is imperative for future health improvements.
- 6Geneticisation reflects the power of geneticists to identify and classify health problems
In sum, Abby Lippman's geneticisation thesis involves a large number of claims concerning the dominance of genetic discourses, changing values and attitudes towards disease and disability, the increasing use of genetic technologies in healthcare, and the powerful role of geneticists in shaping definitions of and responses to disease. Lippman's biography as an epidemiologist, social scientist and health campaigner can be clearly heard in her writing. She is arguing not just against genetic approaches to healthcare, but for an approach which prioritises the social and structural determinants of health. Geneticisation should therefore be situated firmly within a more general critique of public health practices, known as the ‘ecological perspective’ (Lupton 1995).
As Hedgecoe (1998) notes, the idea of geneticisation is based on, and draws on, medicalisation. A number of parallels can be seen between the concepts. Following Strong (1979), one can argue that both ideas are concerned with the individualisation of social/health problems, criticised for obscuring the influence of social processes and the need for social change, and with the expanding jurisdiction of certain professional groups to define and manage problems. While medicalisation talks of the expansion of medicine as a whole, involving boundary work (Gieryn 1983) with other professions and with lay people, geneticisation implies boundary work within medicine, involving the privileging of the status of geneticists and genetics over other sub-disciplines of biomedicine, as well as other professional groups.
In the years since Lippman's original publications, geneticisation has been widely cited and enrolled across a number of disciplines and in connection with a range of phenomena and conditions.2 Just as geneticisation is based on medicalisation, discussions of geneticisation have followed a similar trajectory. These have claimed that the thesis is founded on a narrow and selective empirical base, and that, as initially formulated, geneticisation can only be seen as a negative process (Hedgecoe 1998, 1999, 2001). The discussions have moreover argued that geneticisation should be considered on a number of analytical levels (Hoedemaekers 2001, and ten Have 2000, drawing on Conrad and Schneider 1980), and have taken up the issue of lay agency, suggesting that patients have a more active and productive role in relation to biomedicine than envisaged by Lippman (Novas and Rose 2000).3 Our own approach to geneticisation has been to take the tenets of Lippman's thesis outlined earlier in this section as the agenda for empirical investigation concerning whether and how genetic knowledge and technologies are changing disease concepts, health-related practice, social relations and values.
While geneticisation has generated a great deal of discussion, there have been, to date, relatively few detailed empirical studies that focus on this notion, and even fewer concerned with common complex conditions that are managed outside the genetics clinic. The work of Cox and Starzomski (2003), Hall (2004, 2005) Hedgecoe (2001, 2002) and Lock (2005) suggests that genetic models for common conditions may be emerging at the level of scientific discourse, but raises questions about the current utility of genetic information and technologies in clinical practice and the uptake of genetic ideas by clinicians and patients. In short, the current evidence concerning the geneticisation of common complex conditions is limited and, at best, equivocal.
A genetic vision for CHD
CHD features as a central example in the genetic vision for common complex diseases in both scientific papers and policy documents. For example, in the UK, it is one of the key conditions discussed by both the genetics White Paper (Cm 5791 – II 2003) and the nationally-sponsored UK Biobank (The UK Biobank 2004). Commentaries regarding the potential social and cultural implications of these visions also cite CHD. Lippman (1998), herself, uses heart disease as a key example in her critique of susceptibility testing, public health and geneticisation. Despite the paucity of detailed empirical work in this area, many commentators remain committed to the idea that common conditions are being recast as genetic, including both those who endorse and those who critique geneticisation (see for example Duster 2003, Fullerton 2005, Gannett 1999, Hallowell 1999, Helén 2004, Lemke 2004, Rose 2001, Sherwin 2004). CHD is almost always cited as a key exemplar, along with cancer, schizophrenia, Alzheimer's and diabetes. There is little doubt for these commentators that heart disease is being reconstructed as genetic on a number of levels. In short, the idea of a genetic future for coronary heart disease is firmly embedded within some biomedical, social science/humanities and policy discourses. Nevertheless, there has been little research on the distribution or uptake of genetic discourses in this area.
On the basis of his research in a cardiac clinic and a genetics laboratory, Hall (2004, 2005) argues that there can be no straightforward geneticisation of heart disease because of the contingent and negotiated emergence of genetic knowledge, in this case concerning hypertension, and because of the ‘contesting knowledges’ of cardiologists, nurses, people with heart disease and others. This argument is taken up and developed in the current paper, which focuses specifically on professional discourses. The paper addresses the specific ways in which CHD is being constructed as genetic in the biomedical literature and how widely these genetic models have currency. It aims to contribute to a contemporary account of the place of genetics in models of CHD. In contrast to Hall, our analysis leads us to ask questions about the nature of the concept of geneticisation and the sort of work this concept does.
The paper analyses four commentary articles from the fields of genetics, epidemiology and cardiology. The aim of this analysis is to indicate what aetiological models of CHD are present in the literature, and to show how a genetic model of CHD and alternative models of CHD are constructed. The four articles were selected from a literature search of standard bibliographic databases using the key words and MeSH terms ‘coronary disease, coronary heart disease, aetiology, etiology, genetics, risk factors, and causes’. This was limited to commentary articles, written in English and published between 1999 and 2003. These searches identified a dozen possible articles from which four were selected for analysis. They were chosen because they offered relatively general and accessible accounts. They were also authored by eminent scientists in their respective fields and were published in relatively high-ranking journals.
Journal accounts of an area of science are highly accessible. They are the accounts most likely to be available to and seen by others (Hedgecoe 2001) and provide ‘potent markers of the state of knowledge in a particular field’ (Kerr 2000: 854). One feature of journal accounts is that they tend to minimise ambiguity and controversy (Gilbert and Mulkay 1994, Kerr 2000, Myers 1990a). Because differences and ambiguity are often hidden, it is useful to compare different texts as a way of highlighting the different ways of constructing particular ‘facts’ or issues. This requires sensitivity to the arguments and data that are included and absent, and the understandings that remain implicit (Gilbert and Mulkay 1994, Kerr 2000). The analysis undertaken here focuses on identifying the different strands that contribute to aetiological models of CHD, and the types of arguments and evidence that are enrolled to support them and the aspects that are absent.
The analysis focuses on a limited number of commentary articles and this method draws particularly on the work of Hedgecoe (2001, 2002, 2003a) on the construction of genetic discourses in scientific review papers. Review articles and other commentary pieces play an important and particular role in shaping scientific knowledge, by selecting from existing research papers, and putting these together to synthesise a particular narrative (Myers 1990a, 1990b, 1991). The review draws the reader ‘into the writer's view of what has happened, and by ordering the recent past, suggests what can be done next’ (Myers 1991: 46). They are, therefore, concerned not only with what has happened, but also with shaping the future. Hedgecoe (2003b) shows that commentary articles may well outnumber original research papers in emerging scientific fields. This volume of commentaries would not be warranted if they were only concerned with reporting on current knowledge, but can be understood in terms of their importance in constructing future visions. Following this logic, one would expect to find genetic models of CHD at least in commentary pieces, even if they were absent elsewhere.
It must be noted that Hedgecoe's method and analysis have been criticised by Kerr (2004), who argues that these have produced an ahistorical and inflexible model. This critique should perhaps be understood as relating specifically to Hedgecoe's particular approach and analysis. Nevertheless, in the light of Kerr's critique, it should be acknowledged that the analysis presented here provides a snapshot of a particular field at a particular time. The analysis focuses on the range of models and arguments employed by a limited number of opinion leaders in three important disciplines. These are not intended to be seen as representative or exhaustive of all ideas about CHD, but as providing examples of some of the important ways in which it is currently constructed. The papers are sufficient to support the main arguments of the analysis, i.e. to demonstrate the constructed and contested nature of CHD aetiology and illustrate that there are divergent positions.
The four main texts considered in this analysis are:
- 1Stephens and Humphries (2003) The molecular genetics of cardiovascular disease: clinical implications.This review paper is specifically concerned with genetics and CHD. Professor Steve Humphries is one of the UK's leading scientists in the field of cardiovascular (CV) genetics, Director of a British Heart Foundation funded centre for CV genetics and Chief Executive Officer of London IDEAS (London's genetic knowledge park). Jeffrey Stephens was working as a clinical research fellow at the centre for CV genetics headed by Steve Humphries at the time of publishing. This article was published in the Journal of Internal Medicine, which is ranked 12/105 in the 2005 ISI Journal Citation Reports, category medicine, general and internal.
- 2Beaglehole and Magnus (2002a) The search for new risk factors for coronary heart disease: occupational therapy for epidemiologists?This is a review and discussion piece concerning CHD risk factors and public health strategy. Robert Beaglehole is a well-known public health physician, currently working as an adviser to the World Health Organisation. Paul Magnus, perhaps less well known internationally, is the medical advisor to the Australian Institute of Health and Welfare. The analysis also looks at the ensuing responses and discussion this article prompted. This was unusual in that disagreements are made explicit in the texts. It thereby casts into sharp relief the ambiguities involved in the construction of medical knowledge. These articles were published in the International Journal of Epidemiology, which is ranked 6/99 in the category public, environmental and occupational health.
- 3Lenfant (1999) Conquering cardiovascular disease: progress and promise; and
- 4Beller (2001) Coronary heart disease in the first 30 years of the 21st century: challenges and opportunities.These authors were respectively Director of the American National Heart, Lung and Blood Institute and President of the American College of Cardiology at the time of writing these pieces. Lenfant's editorial was written to mark the 50th anniversary of the Institute and Beller's lecture to mark the turn of a new century. Both these texts very clearly provide the authors with a space to create an account of the current field and envisage the future prospects. Lenfant's article was published in the Journal of the American Medical Association, ranked 3/105 in the category medicine, general and internal. Beller's article was published in Circulation, which is ranked 1/72 in the category cardiac and cardiovascular systems.
A genetic model of CHD
Stephens and Humphries’ (2003) paper is concerned with the possible effect of functional polymorphisms4 in the aetiology of CHD. Their paper discusses three examples of these and the possible interaction, at the molecular level, of these three genes with environmental factors, focusing particularly on smoking. The paper demonstrates a familiar narrative about a genetic future for understanding and managing cardiovascular disease. CHD is presented as a polygenic, multifactorial condition in which genetic risk results from the cumulative effect of variations in several different genes. The overall message is that understanding the role of genetic polymorphisms is key to understanding the molecular and cellular pathology of CHD, which in turn may be important for the development of new therapies, diagnostic tests and prevention strategies. At the same time the authors acknowledge the importance of the more established risk factors and practice and a large part of the paper is devoted to gene-environment interactions. These are interactions where the cumulative risk is greater than the individual risks of each factor added together. The main objective of genetic research into CHD is cited as to ‘provide a complete risk assessment, complementing the well-studied traditional clinical and biochemical risk factors’ (2003: 125–6, emphasis added). If one understands review articles as a space where futures are envisaged and a case is built for claims on resources, one might see the framing of this introduction as a way of helping to carve out a legitimate place for genetics in the field of CHD, alongside the established disciplines such as clinical pathology and cardiology.
The first part of the paper is devoted to establishing the importance of genetic factors in the aetiology of CHD. Their key argument is that:
functional gene polymorphisms account for much of the biological diversity in homeostatic systems. In their absence all humans would respond in an identical manner to an environmental challenge, and the risk of developing disease would be directly proportional to the environmental stimulus. We know that this is not the case. For example, some individuals exposed to cigarette smoke with an otherwise identical risk factor profile will go on to develop CHD, whilst others will not. Therefore, the well accepted view is that CHD is a multifactorial disorder, with both environment and genetic factors contributing (2003: 120–1, emphasis added).
This argument hinges on the idea that it is possible to identify people with otherwise identical risk factor profiles, which implies that all the risk factors for CHD are already clearly established and that they can be controlled for. The use of the phrase well accepted view seems to close off any scope for discussion. It is, however, not an argument that is universally accepted by biomedical scientists, as the analysis provided later demonstrates.
In the paper it is particularly striking how genetic differences, environmental risk factors, and the relationship between environment and genes are constructed. Stephens and Humphries appear to use the terms ‘mutation’ and ‘polymorphism’ interchangeably with the same variations referred to as both ‘minor mutations’ and as ‘functional gene polymorphisms’. Variations in the apolipoprotein-E (ApoE) gene are discussed. This gene codes for a protein involved in lipid transport.5 There are three ApoE variants, E2, E3 and E4, and their frequencies are given as 8 per cent, 77 per cent and 15 per cent respectively in white populations. Through the discussion E3, the most common variant, emerges as the normal gene. The other variants are referred to as ‘single common mutations with modest impact’ (2003: 121). Since more than 40 per cent of the population have at least one copy of the E2 or E4 variants, it could be argued that this effectively constructs a large proportion of the population as abnormal in some way. Kerr (2000) describes a similar ‘re-classification’ of polymorphisms into mutations as one of the reductionist processes by which researchers built their case for an association between cystic fibrosis and a form of male infertility.
An interesting conception of environmental risks also emerges in the paper. Environmental challenges are said to include: ‘diet, male sex, diabetes, obesity and cigarette smoking’ (Stephens and Humphries 2003: 121). It is not clear what the authors have in mind in their reference to diabetes. They may see it as a proxy for ‘lifestyle’ factors such as inappropriate diet and lack of exercise which are associated with obesity and type 2 diabetes. It could, however, also be read as a physiological or bodily state concerning glucose metabolism. The inclusion of male sex, i.e. being male, certainly suggests that the environment envisaged by the authors is the biological environment of the genes or of the organs rather than the physical or social environment of the person. The authors also link environmental risks and lifestyle choices, seeming to conflate the two concepts: ‘individuals adopt a different position on the environmental spectrum of risk by the lifestyle choices they make (e.g. smoking). However, although environmental risk factors are modifiable the genetic factors are not’ (2003: 124). The excerpt shows that, in this paper, one of the defining attributes of the environment is that it can be changed. It is ironic, then, that, male sex is included as an environmental factor. This rather obvious contradiction highlights the limited conceptualisation of environment employed.
It is not clear in the model outlined in this paper whether environment is envisaged as being able to play a role on its own in the absence of genetic predisposition. CHD is referred to as a ‘polygenic’ disease. The only exceptions that the authors recognise are cases where the disease is monogenic, involving a single rather than multiple genes. The implication is that genetic variations are always involved in CHD in some way. The paper also argues that ‘CHD is a complex condition resulting from numerous gene-gene and gene environment interactions’ (2003: 120). This supports the idea that genes are a necessary part of CHD aetiology. On the other hand, the authors go on to comment that CHD may occur ‘as a result of failure at the genetic level (e.g. gene transcription) or, due to an environmental exposure (e.g. smoking) or due to an imbalance between the two’, implying that environment alone may cause CHD in some cases (2003: 120).
Taken as a whole, this paper appears subtly to privilege genetic factors. There are parallels here with Hedgecoe's (2001) analysis of biomedical discourses about schizophrenia. Hedgecoe suggests that these discourses under-specify and downplay the environment and prioritise genetic factors to the extent that a ‘genetic baseline’ is seen as necessary for the causation of the disease. Hall's (2004) ethnography of a laboratory conducting research on the genetics of hypertension adds another layer to this discussion. His work suggests that essentialism and uncertainty are strategically enrolled at different points in the funding cycle. Like Hedgecoe, however, Hall argues that even though researchers sometimes acknowledged the role of social and environmental influences, the overall picture presented supported the idea of a genetic baseline in the case of heart disease. In the paper analysed here, while the environment is clearly under-specified, the conclusions regarding the notion of a ‘genetic baseline’ for CHD are more tentative.
Nevertheless, overall, Stephens and Humphries’ account of the causes of CHD does seem to attribute a large role to genetic variation between people, creating new categories of biological difference. The interchangeable use of the terms variations, polymorphisms and mutations tends to pathologise some of these new categories of people. It also presents a rather incomplete view of environmental factors in which the social and physical context are absent. Further, their discussion suggests that new genetic knowledge is imperative for progress in understanding CHD and will be useful for treating the disease. In other words, in terms of Lippman's definition, this is a thoroughly geneticised view of CHD. It is, however, not the only available model of CHD, as the following sections clearly demonstrate.
Epidemiology fights back
This section considers the article by Beaglehole and Magnus (2002a), four commentaries that followed on from this article (Greenland et al. 2002, Law 2002, Marmot 2002, Nieto 2002) and a response by the original authors (Beaglehole and Magnus 2002b). This dialogue provides an entirely different perspective on the aetiology of CHD and demonstrates that genetics is just one of a number of emerging and contested areas of interest.
Beaglehole and Magnus's (2002a) basic argument is that the major risk factors for CHD have already been established. These are high blood cholesterol, high blood pressure, cigarette smoking and physical inactivity. Furthermore, they attribute these to economic, social and cultural factors. They argue that:
These major causes have a close and precisely defined (proximal) relationship to the CHD epidemics and are well established scientifically (2002a: 1118).
It is notable that what is well established scientifically, in other words, what every one knows about CHD in Beaglehole and Magnus's world already differs from what everyone knows about CHD, or the well accepted view, in Stephens and Humphries’ world.
Beaglehole and Magnus argue that the search for new risk factors is partly based on an underplaying of the role of the main factors they list, as a way of justifying new research. They argue that there is a widely accepted ‘myth’ that the established CHD risk factors explain only half or less of the occurrence of CHD, whereas in their view the major risk factors explain at least 75 per cent of new cases of CHD.
They go on to discuss the new risk factors, which fall into six main groups:
- • Thrombotic factors and the effect of biochemical markers;
- • The role of inflammation and infectious agents;
- • The influence of early life exposure;
- • The contribution of multiple genes;
- • Oestrogen deficiency;
- • The role of the psychosocial environment (2002a: 1119).
The authors evaluate these factors in terms of their potential contribution to prevention of CHD and the degree to which they contribute to explaining ‘CHD epidemics’. They are particularly critical of genetic research and better disposed towards some of the other factors. In the end, however, all the factors are variously dismissed, due to lack of conclusive evidence or because their potential contribution to population health is doubted. The majority of funding, they argue, should be directed to prevention strategies at the primary and population levels, including structural changes concerning food and tobacco retailing, and access to physical activity. In an argument echoing Lippman's, they conclude that preoccupation with identifying high risk groups and individuals reflects the dominant health paradigm, which, they argue:
supports an individualistic approach to health improvement and ignores the wider social and economic determinants of the health of populations (2002a: 1121).
It is notable that, in contrast to Stephens and Humphries, environmental factors are framed in terms of the cultural, economic and social factors that lead to CHD risk, rather than as ‘life-style choices’.
Beaglehole and Magnus's argument is essentially that there is a gap between current knowledge and practice regarding population levels of the established risk factors and there should be more focus on prevention strategies at this level. None of the four commentaries that follow the article take issue with these main arguments. What is at stake is more the strength or weakness of Beaglehole and Magnus's arguments concerning the new or emerging risk factors and their notion of a ‘public health test’ of a risk factor. The commentaries make three main types of argument: they propose alternative aims for epidemiological research, dispute the evidence proposed by Beaglehole and Magnus and provide counter evidence and arguments to support the ‘new’ risk factors.
While Beaglehole and Magnus are solely interested in population levels of CHD, the commentators propose that there are other legitimate aims for epidemiological researchers. Marmot (2002), for example, is particularly interested in why rates for different groups of people vary so greatly. He asks why ‘among people more or less equally exposed, there remain such marked differences in the rate of occurrence of CHD’ (2002: 1124). His own work is concerned with explaining the social gradient in CHD and has particularly focused on psychosocial factors. It is notable that Stephens and Humphries employed the idea that only certain individuals with a given level of the known risk factors go on to get CHD as an unequivocal argument for genetic influences, implying that there was no other possible explanation. Yet, the same argument can be employed to support research into other risk factors.
Commentators are critical of the idea that 75 per cent of CHD is explained by the established risk factors. They also question the evidence provided by Beaglehole and Magnus in their evaluation of the emerging risk factors, criticising both individual papers cited, for example, pointing out the limitations of a particular meta-analysis concerned with infection with Chlamydia pnemoniae, and the types of evidence drawn on in general, for example, a reliance on observational evidence and a failure to consider pathology, laboratory and experimental evidence. These discussions make clear that there is a range of proposed risk factors for CHD and they bring to the surface the uncertainty and contingency of scientific claims. They also illustrate that the same kinds of arguments can be enrolled to support different types of models of CHD. What is notable is that no defence of multiple gene influences is made. In these discussions, in any case, genetic explanations appear to be low on the agenda.
These discussions also illustrate a point made in the introduction to this paper, that Lippman's concerns form part of a wider critique concerning individualistic approaches to health and a lack of attention to the wider determinants of health and illness. It is notable that while Lippman's criticism is directed towards geneticists and their allies, both Beaglehole and Magnus (2002b) and Marmot (2002) make antagonistic comments about the role of cardiologists in individualising prevention strategies for CHD:
without strong epidemiological input and leadership, cardiologists will continue to dominate the prevention debate and individual approaches will remain the priority (Beaglehole and Magnus 2002b: 1134).
In the next section it is suggested that although clinicians may tend to focus more on individual-level prevention, this does not mean a wholesale adoption of genetic discourses.
The clinicians’ view
This section considers papers by two clinicians, Lenfant and Beller. Lenfant's (1999) editorial piece entitled ‘Conquering cardiovascular disease: progress and promise’ sets out to provide an overview of recent progress and issues in cardiovascular disease research. He makes three main arguments in relation to coronary heart disease: (1) there is still work to be done concerning the established risk factors, (2) molecular biology and molecular genetics may offer major improvements in the field of cardiovascular disease, and (3) there is a need for better application of the established interventions. In other words, there is a gap between knowledge and practice.
Lenfant is optimistic about the role of genetics:
On the eve of the next millennium it is safe to predict that the importance and application of molecular genetics and functional genomics will play major roles in the further improvement of cardiovascular health (1999: 2069).
Although he makes reference to genetic polymorphisms, equally he includes discussion of developments in molecular biology concerned with the role of various molecules and cellular processes in the development and stability of atherosclerotic plaques and the function of the arterial lining. It is important to note that such developments may be concerned with understanding the molecular processes that lead to CHD, in general, and with finding new ways of assessing the risk of and progress of the disease, based on this knowledge. It must be recognised that such research is not necessarily related to identifying genetic differences between people. In other words, interest in molecular biological processes should not be conflated with genetic models that are based on inherited differences between people.
A major part of Lenfant's argument is that there is a problem translating research results into lifestyle changes, public health interventions and clinical practice, and he highlights this by discussing evidence concerning poor use of drugs for blood pressure and poor ‘compliance’ with blood-pressure treatments. He concludes that although molecular biology and genetics offer exciting and useful prospects, doctors should not forget about the ‘more mundane’, but effective methods like lowering blood pressure, decreasing obesity and physical inactivity and the appropriate use of established therapies such as β-blockers and aspirin:
The real challenge for the new millennium may indeed be to strike an appropriate balance between the pursuit of exciting new knowledge and the full application of strategies that already are known to be extremely effective, but considerably underused (1999: 2070).
Beller's (2001) review, entitled ‘Coronary heart disease in the first 30 years of the 21st century: challenges and opportunities’, would seem to provide a perfect opportunity to focus on predicted developments in and applications of cutting edge technologies such as genetics. In Beller's hands, however, the major challenges for CHD prevention and treatment are framed in more prosaic terms. Discussion focuses on three main factors: (1) an increasing proportion of older people in the population, (2) an ‘epidemic’ of type 2 diabetes, and, linked to this, (3) an ‘obesity epidemic’. Beller's main response to these challenges is to draw on a discourse of a gap between knowledge and practice. He suggests that older people are already under-treated, arguing that they are not benefiting enough from ‘proven diagnostic and therapeutic strategies’ (2001: 2429, emphasis added). The paper relates diabetes and obesity in part to genetic factors. Nevertheless, the main thrust of the discussion about these conditions is concerned with physical inactivity and poor diet, which are related, amongst other things, to the consumption of ‘fast foods’, increased television watching and computer games and a lack of physical education programmes in schools. Beller calls for greater emphasis on prevention in these areas, stressing the responsibility of cardiologists, parents and educators. In contrast to Beaglehole and Magnus, and with the exception of education policy, these are largely taken as individual problems requiring individual actions.
For the future, an array of technological developments, both biological and mechanical, in all areas of prevention and treatment are predicted. Beller mentions the prospect of genetic screening to identify people at risk of diabetes and CHD, but appears to link this solely to prevention for people at high risk of early CHD. In other words, this is a fairly limited role for genetic developments, rather than the more global role suggested by Stephens and Humphries (2003). In conclusion, he concurs with Lenfant and urges us not to forget the ‘less costly, low-tech interventions [that] have already proven effective in preventing CHD and its complications’ (Beller 2001: 2434).
In sum, Lenfant and Beller display some enthusiasm for developments in genetics. Again, however, this could certainly not be thought of as the dominant theme in their accounts of the future. Their message is that health improvements may result from developments in genetics but, more importantly, could be accrued through better implementation of a raft of established low-tech (or even no-tech) interventions, i.e. through better application of what is already known.
The analysis demonstrates that there is not one biomedical model of CHD in the literature, but a number of streams of discourse running in parallel. These different ways of understanding CHD prioritise different causes, different ways of preventing and managing CHD, and call into being different futures for the field of CHD. In short, there is significant heterogeneity in biomedical discourses about CHD in the literature. There certainly is a genetic model of CHD that has many of the elements of the geneticised discourse proposed by Lippman. The account of CHD provided by Stephens and Humphries privileges a genetic explanation while detracting from other causes, and creates new categories of biological normality and abnormality. It also suggests that new genetic knowledge is imperative for increased understanding of CHD and may improve treatment of the disease. At the same time, the models presented by the public health physicians, Beaglehole and Magnus, and the clinicians continue to focus on the established risk factors, arguing that there is scope for major improvements if only better use were made of current knowledge and interventions. The pivotal difference in their models of CHD was their emphases on population/strategic versus individual-level interventions. Other epidemiologists were concerned with explaining variations in the occurrence of CHD that cannot be accounted for by the established risk factors (although how much is already explained is contested). Their discussions included a number of elements that may contribute to the aetiology of CHD, including infections, clotting processes, early life experiences, oestrogen levels and multiple gene influences. In other words, these models involve a range of possible risk factors, of which genetics is just one.
There are currently a number of ways of understanding the role of genetics. It can be seen as providing a unifying framework, in which genetic variations are integral to and necessary for understanding disease pathways, predicting susceptibility and explaining all cases of CHD. It can also be seen as an explanation for a circumscribed and limited subset of early or ‘premature’ CHD. Alternatively, it may mean little more than a general and more long-standing recognition that CHD can run in families. Genetics certainly did not represent the dominant discourse, as its place was highly circumscribed in all but the account given by molecular geneticists. Therefore, like Hall (2004, 2005), we find that, at this rhetorical level, there is no straightforward geneticisation of CHD – genetics was not necessarily seen as central or essential to the future of CHD prevention and care.
Lippman's claims about the power of geneticists to condition how a host of diseases are viewed seems, at the least, to be overstated. The analysis highlights that there are many disciplines with an interest in CHD and demonstrates that there are a number of different ways of constructing the condition, thus reinforcing Hall's (2004) idea of ‘contesting knowleges’ in relation to CHD. These various constructions should come as no surprise. It reflects the different concerns, objectives and professional practices of the different disciplines, including the types of technologies and evidence they draw on. This is not to say that there are clear distinctions between different disciplines or that each discipline is homogenous (Bucher and Strauss 1961), as the arguments between the epidemiologists illustrate. Nevertheless, CHD must be understood as an object shared across a number of domains of biomedicine. In this sense, it can be understood as a boundary object (Star and Griesemer 1989), in other words as a single object which has different meanings in different domains, but remains recognisable across these domains. More recent STS writing would even suggest that it is a multiple object that is enacted differently through specific local practices (Law and Singleton 2005, Mol 2002). Understanding CHD in this way means that one can only talk of certain models dominating in particular and local sites, such as certain literatures, clinical encounters, laboratories, or areas of policy and funding. In other words, CHD has local meanings and ones that are specific to certain professional discourses and practices.
The commentary pieces analysed here can be understood as setting out specific visions of CHD prevention and care that are associated with certain professional groups and their working practices. As such, the articles can be seen as a form of boundary work (Gieryn 1983), which involves elaborating and defending the strategies particular professional groups draw on to demarcate their own discipline from others, and which establish their own expertise, authority and claims on resources in a particular field. This was most visible here in the articles written by epidemiologists in their claims for control over prevention strategies for CHD and admonition of cardiologists for promoting individual-level approaches to prevention. There were hints of boundary work in the other papers too. In the case of molecular genetics, Stephen and Humphries’ (2003) framing of genetic knowledge as complementary to established clinical knowledge helps to give genetics a legitimate place at the CHD table, alongside more established disciplines. One might also suggest that the clinicians, in arguing for the better application of current knowledge regarding such issues as the treatment of raised blood pressure, are working to maintain their central role in the management of CHD. In sum, there are multiple meanings of CHD and these are embedded in particular professional domains. Alternative claims about the aetiology and management of the disease are central to struggles over authority between the different professional groups.
In contrast to an understanding of CHD as an object that is interpretively flexible, or even multiple, geneticisation makes a universal claim that involves a dominant way of thinking about and ‘doing’ a particular disease across all domains and sites. So, is it possible to imagine any circumstances under which a condition could become geneticised? The first is where clinical genetics has an effective monopoly over the aetiology, diagnosis and treatment of a disease. This appears to be the case with classic monogenic conditions, such as cystic fibrosis (CF), although, even here, there are questions about the utility of genetic information and technologies in clinical practice (Featherstone et al. 2005, Hedgecoe 2003a, Kerr 2000, Shaw 2003). In any case, this is unlikely to occur with the majority of established common disease categories. The second is where new genetic explanations and technologies become the dominant way in which all other professional groups involved in understanding and managing a particular disease carry out their practice. Again, this seems unlikely to occur, at least in the foreseeable future, as it would require established professional groups to abandon entrenched ways of thinking and working that to a large extent define their areas of expertise.
This conclusion, therefore, raises a more general question about the nature of the geneticisation thesis. Once one starts to think about the commentary papers in terms of the work they do in constructing and maintaining particular boundaries, it becomes possible to consider the geneticisation thesis itself, as a form of boundary work. Just as epidemiologists argue that they should retain jurisdiction over CHD prevention vis-à-vis cardiologists, Lippman, as an epidemiologist and social scientist, is involved in boundary work concerning geneticists, warning against their colonisation of areas of health and illness which she argues would be better seen in other ways. Strong (1979) made a similar argument about the medicalisation thesis, although he frames this in more inflammatory terms (labelling the thesis a form of ‘sociological imperialism’). The current findings reinforce the argument that geneticisation should be seen as part of a wider project concerned with lobbying for greater emphasis on population/strategic-level rather than individual-level interventions in the field of health. This perspective may provide a more useful analysis of the concept than a literal reading, which, as shown here, is highly unlikely ever to occur.
While recognising that the analysis presented here focuses on just one level of data and a small sample of papers, it strongly suggests that there are a number of significantly different ways of understanding CHD and a degree of heterogeneity that would not be anticipated by proponents of geneticisation. As a totalising claim, geneticisation is at odds with the diffuse and distributed nature of biomedical knowledge and practices. Social and humanities analysts, therefore, need to be cognisant of the work they do when they assume, rather than question whether, genetic discourses and practices are spreading. The paper contributes to a growing perception that researchers should be wary of investing genetics with a high degree of agency. In this sense, it may be valuable to question the extent to which social science analysis, itself, has become geneticised.
Many thanks to Alison Kraft, Elizabeth Murphy, Robert Dingwall, Brigitte Nerlich and the two anonymous referees for their helpful comments on earlier drafts of this paper. The research was supported by the Economic and Social Research Council, Postgraduate Studentship PTA030200200082.
Commentaries include reviews, editorials and discussion pieces. These discuss or draw together published research. They are distinguished from original research papers, which present new, previously unpublished research.
The Web of Knowledge lists 142 papers that cite either Lippman (1991) or Lippman (1992) – search undertaken on 30/3/06.
See Strong (1979) and Lupton (1997) for parallel arguments about medicalisation.
Genetic polymorphism means gene variations. Some variations lead to variations in the gene product and may be referred to as ‘functional gene polymorphisms’. These may or may not be thought to be disease causing or pathogenic. Mutation also means variation in a gene. It can mean merely a random change in the DNA, but in medical papers it usually denotes a pathogenic change, and often means an inherited pathogenic change. There is no precise distinction between the terms polymorphism and mutation. Use of polymorphism is often associated with more prevalent variations. It tends to be used only where a variant occurs with a frequency of at least one per cent. More importantly, in biomedical papers, polymorphism tends to be used for benign variations and mutation for pathogenic changes.
Lipids include cholesterol and fat particles such as triglycerides. They are transported around the body in the form of lipoproteins, such as low-density lipoprotein (LDL) and high-density lipoprotein (HDL). The apolipoproteins, including ApoE, constitute the protein component of lipoproteins.