Introduction

Obesity is considered one of the most pressing health problems of our time, especially in developed countries (Bell et al, 2005), where it is treated as a public health disaster (Naser et al, 2006; Hobbs, 2008) of purportedly epidemic proportions (World Health Organization, 2010). Despite receiving incredible amounts of resources, public attention and energy, consensus with respect to the definition, causes or solutions is lacking. Two explanations and consequent interventions are advanced by governments, health authorities, healthcare practitioners, scientists, and public interest groups. First, especially prominent in media (Saguy and Almeling, 2008) and public health programmes, obesity is a social problem (Boero, 2007) spurred by inadequate education about healthful eating and exercising, the normalisation of poor food and activity choices, an ‘obesogenic environment’ (Swinburn et al, 1999), and individuals’ lack of self-control, will and self-worth. Second, biophysiological pathways including ‘fat genes’ (Throsby, 2007), ‘obese genotypes’ (Bell et al, 2005), endocrine abnormalities (Bjorntorp, 1995) or viruses (Atkinson, 2008) lead to the unhealthy accumulation of fat. Both explanations presume that accumulation of fat on bodies is pathological, an assumption challenged by the ‘fat acceptance’ (FA) movement, which promotes ‘health at every size’ (see Robison, 2005).Footnote 1 This fat consciousness challenges the pathologisation of fat, emphasises well-being over weight loss, rejects moral panic surrounding fat, questions dominant representations of the mutual exclusiveness of beauty, health, fitness and fat, celebrates diverse bodies, demands responsive healthcare, and critiques fat discrimination. This politicisation of fat is taken up in a growing field of research called fat studies (for introductions see Rothblum and Solovay, 2009; Cooper, 2010) or critical obesity research (Colls and Evans, 2009), which approaches obesity as ‘as a socially constructed and contested “problem” rather than as an unproblematic category, disease or aetiology’ (Keenan and Stapleton, 2010, p. 370). These perspectives comprise an expansive tug-of-war over obesity with many ropes and differing numbers and strength of pullers, at the centre of which are those individuals deemed obese by biomedicine, governments, publics, and/or themselves whose quality of life, life opportunities and health lie in the balance.

These vastly disparate perspectives on fat suggest that obesity is, to use the language of Sarah Whatmore (2009), a ‘knowledge controversy’: when competing claims and bases for expertise, especially those of publics versus dominant scientific or governmental ones, mar actionable consensus on a pressing issue. For Whatmore, knowledge controversies can be addressed by meaningfully accounting for, redistributing and integrating expertise broadly defined. In this article, I suggest that the Barker hypothesis, dubbed in recognition of the foundational work of epidemiologist David Barker (Paneth and Susser, 1995),Footnote 2 can support the politicisation and redistribution of obesity expertise. Now an established theory also referred to as the ‘developmental origins of adult disease’ (Barker, 2004), the ‘foetal origins hypothesis’ or the ‘developmental origins of health and disease’ (DOHaD),Footnote 3 a vast scientific literature confirms its basic argument: many diseases throughout the life course can be traced to the conditions for development of the foetus in the maternal womb (see Paneth and Susser, 1995; Eriksson et al, 2000; Godfrey and Barker, 2001; Lumbers et al, 2001; Barker, 2004; Bateson et al, 2004; Hall, 2007; Barker et al, 2010a, 2010b, 2010c; Eriksson et al, 2010). These events and conditions are determined by the commingling of temporally and spatially complex biological and social processes captured in the two key concepts of developmental plasticity and foetal programming. As such, the Barker hypothesis does not support solely reductionist, biophysiological paradigms of health and disease. Instead, in principle it evinces complex understandings that span biology, social positionality, place and generation.

The hypothesis should therefore be seen as part of a larger paradigm shift towards complexity (Thrift, 1999; Gilbert and Sarkar, 2000; Urry, 2003, 2005; Capra, 2005) and globality (Clark, 2005, 2011; Hird, 2010), making it significant for science studies; biopolitics of health, illness and justice; behavioural and biological theories of development; and, as is the particular focus of this article, transdisciplinary studies of health and disease. Defined as a research practice in which ‘boundaries between and beyond disciplines are transcended and knowledge and perspectives from different scientific disciplines as well as non-scientific sources are integrated’ (Flinterman et al, 2001), transdisciplinarity has gained currency as part of a growing recognition that responses to wide-ranging health concerns must draw on many forms of data and analytic resources and be multi-pronged and multidimensional. Inasmuch as dominant approaches to obesity have not harboured such a recognition, they have contributed to the marginalisation of people deemed obese and generated ineffective or harmful interventions. I argue that the hypothesis prompts consideration beyond the conventional bounds of epidemiology to new sources of understanding and new sites of action that may influence and support movements concerned with body politics and justice for fat people.Footnote 4 Finally, I ask why the potential of Barker hypothesis has not actualised in the form of the transdisciplinary projects it begs. I point to literature on the potential for injustice engendered by the Barker hypothesis, and suggest that these critiques reveal the very necessity for transdisciplinary collaboration on obesity in the first place.

The Problem of/with Obesity

Lee F. Monaghan's (2005) essay on the contours and controversies of the ‘obesity debate’ powerfully reminds that any discussion of fatness must not reify obesity as a transparent disease or problem. Obesity is so often framed as an issue of ‘calories in versus calories out’ that the characterisation has become dogmatic in light of competing explanations. This explanation is intimately tied to the ‘war on obesity’, in which, as Herndon (2005, pp. 129–130) describes, advocates

imagine themselves engaged in a battle for our nation's health because they believe obesity poses a significant public health threat …. What many doctors, public health officials and concerned journalists writing in support of the war against obesity fail to recognize, however, is that a war against obesity also means a war against fat people …. Obesity is not a pathogen, not free floating, and never a virus that attacks a helpless and innocent victim. Instead, obesity is virtually always typecast as a condition brought on oneself. A war against obesity, then, cannot be a war against a faceless pathogen. Instead, obesity is a condition of human causation and therefore necessitates a war against the group of people participating in the volitional behaviours that cause it.

In locating responsibility in individuals, communities or society at large, obesity has become a moral issue (Campos et al, 2006) and even a religious one (Herndon, 2008). Stereotypes that fat people are uneducated, weak-willed and immoral (Campos et al, 2006; Boero, 2007; Throsby, 2007) in ‘choosing’ to be fat or denying their ‘real’ eating and exercising habits (Herndon, 2008) occlude the ‘epidemic’ in moral panic (Fraser et al, 2010; Rail et al, 2010). The war on obesity therefore attaches significant social stigma to fatness. Studies have exposed the systematic marginalisation of fat people in workplaces, healthcare, education and other areas (Pomeranz, 2008), which can damage their personal health and well-being (Rothblum et al, 1990), life opportunities and quality of life. Importantly, the war on obesity has yet to be won or even make minor gains. Linda Bacon, FA advocate, writes:

we’ve lost the war on obesity. Fighting fat hasn’t made the fat go away. And being thinner, even if we knew how to successfully accomplish it, will not necessarily make us healthier or happier. The war on obesity has taken its toll. Extensive ‘collateral damage’ has resulted: food and body preoccupation, self-hatred, eating disorders, discrimination, poor health …. Few of us are at peace with our bodies, whether because we’re fat or because we fear becoming fat.Footnote 5

Fat people frequently echo the same predicament: despite partaking of the typical interventions of dieting, exercise, diet pills, prescription medication or gastric surgery, many ‘obese’ and ‘overweight’ people are not able to reach or maintain a weight within the ‘normal’ range of body mass index (BMI) or maintain what they see as a healthy body and/or a happy life while also maintaining this ideal (Throsby, 2007).

An increasingly organised and vocal group of individuals and organisations, such as the National Association to Advance Fat Acceptance,Footnote 6 challenge the dogma that only by losing weight can fat people claim health. Some FA proponents acknowledge ‘unhealthy’ practices, but point to the hypocrisy that fat people are held more accountable than ‘normal’ weight people who may, in fact, partake in the same practices that are a part of everyday life (such as eating readily available and cheap junk food, consuming alcohol or commuting by car). Heshka and Allison (2001, pp. 1402–1403) note that

Many obese persons are competent, functioning members of society … [T]hese persons [do not] necessarily subjectively consider themselves impaired, except perhaps insofar as they feel themselves victims of social discrimination. They might fail to meet some arbitrary standard of physical fitness (e.g. climbing stairs, running) but such a standard would also exceed the capability of many non-obese but sedentary individuals.

FA advocates also question the need for and efficacy of medical and commercial interventions that may produce ‘side effects’ that are worse than the health problems of obesity.Footnote 7 Claims about the economic costs of obesity – identified as between 2 and 7 per cent of total health costs in developed countries (Kopelman, 2000, p. 636) – are made in order to prove that capital-intensive interventions are necessary, but these claims must also be understood to equally reflect a desire to make money (Guthman and DuPuis, 2006; Oliver, 2006).

Judgements of the unhealthiness of fat people often come from merely visual assessments (see Boero, 2007; Jutel and Buetow, 2007), determined by norms of beauty and the ‘aesthetics of health’ (Spitzack, 1990; Jutel and Buetow, 2007). Yet because visual ideals are diverse between cultures, beauty does not equate wellness, and beauty is emphasised ‘often at the expense of truly health-protecting and health-promoting behaviours’ (Jutel and Buetow, 2007, p. 422), visual assessments are wanting indicators of healthiness. Normative standards of beauty and aesthetics omit obese people (Jutel and Buetow, 2007; Snider, 2012) and even dehumanise them, the ‘headless fatty’ phenomenon seen in media, in which ‘a fat individual [is] usually shown as a static or walking torso, highlighting the belly, thighs, and buttocks’ (Snider, 2010, p. 174; Snider, 2012), an especially apparent example.

FA advocates are especially critical of the belief that fat people are ‘to blame’ for their adiposity and seek out scientific and other kinds of knowledge that complicate or disprove this assumption. As such, science and medicine play a more ambivalent role in FA than a cursory look at the movement might suggest. For example, FA proponents leverage knowledge of biological causes to counter-balance the emphasis on lifestyle and personal failure of fat people (Saguy and Riley, 2005), enabling them to make disability, insurance or other kinds of claims. Likewise, FA and critical obesity studies draw on inadequacies elaborated by biomedicine to critique BMI, the standard ‘proxy’ (Franzosi, 2006) for roughly determining whether an individual is obese (defined as BMI greater than 30; Friedman, 2009). BMI figures prominently in the ‘essential truth’ of fatness and ‘drives anti-obesity policy [and] governmental policy that seeks to prevent, treat and reduce rates of obesity, despite its inadequacies’ (Evans and Colls, 2009, p. 1052). As often argued, BMI is limited in dealing with very muscular individuals, does not distinguish between lean and fat mass, and may lead to overestimation or underestimation of the health risks associated with body fat for young adults, those over 65 (Health Canada, 2006) and children.Footnote 8 BMI has been further critiqued as ‘alarmist’ (Cooper, 2010, p. 1021) in normalising and pathologising certain body types, presenting risk from accumulated fat as a continuum from low to high BMI and assuming fat is unhealthy. Even if a person's actual fat mass is known by more accurate techniques, Oliver (2006, p. 612) argues that ‘there is no clear evidence about what level or even how, exactly, adipose tissue is harmful to our health’. The pathological significance of fat is a proxy for risk for comorbidities. For example, Kopelman (2000, p. 636, emphasis added) relays a standard definition of obesity as ‘disease in which excess body fat has accumulated such that health may be adversely affected’. Heshka and Allison (2001) critique this, arguing that risk is not an acceptable definition of disease. Many risk factors for the same comorbidities such as age are not similarly pathologised. Definitions of obesity are generally tautological (Heshka and Allison, 2001, p. 1402) in that, as Jutel (2005, p. 114) remarks, ‘overweight is now identified as a disease entity rather than a statistical observation’.

Science and medicine participate integrally in the marginalisation of fat people in the war on obesity. However, resources for critical examination of obesity are also found in science. This article contributes to the mobilisation of scientific knowledge for critical understandings of obesity by exploring the Barker hypothesis.

The Barker Hypothesis

The Barker hypothesis is primarily concerned with development, the course of growth of an organism in utero. Development originates in the profound potential of the foetus in conversation with its milieu. During its time in utero, many aspects of foetal morphology and physiology have myriad possible iterations, such as the number of nephrons comprising the kidney or the bias of its metabolism. Such plasticity allows the foetus to ‘prepare’ itself for the milieu outside the womb that it indirectly ascertains from its mother's physiology, and makes the most of the nutrient and other resources made available by the placenta. Plasticity, then, is the capacity for development as response to experience; as Barker remarked in a 2010 Lecture, ‘development requires experience’. Plasticity is slowly replaced with rigidity until the forms attained will be more or less persist, making development a unique and critical time in which foundations for the life course are set. According to the Barker hypothesis, this has major implications for later experiences of health and disease and, as I argue, fat embodiment.

One cannot attend a conference or meeting in any specialty of reproductive sciences and not hear Barker or DOHaD uttered. Many sessions at national and international conferences around the world focus on the hypothesis. The International Society for DOHaDFootnote 9 hosts its own annual conference, and an entire journal, Journal of Developmental Origins of Health and Disease,Footnote 10 is dedicated to publishing research about early origins. The Barker hypothesis has had a major impact on the study of reproduction and disease, but, perhaps surprisingly, it has also become integral to much of epidemiology and basic science in the study of diabetes, cardiovascular disease, obesity and osteoporosis, and is even increasingly important in studies of cancer, mental illness and many more diverse conditions. In a recent issue of Time, the Barker hypothesis was said to prompt a ‘revolutionary shift’ (Paul, 2010, para 4), indicating that it is also making its way into public and popular consciousness.

Barker and collaborators began work in the 1980s examining historical data on infant mortality and adult cardiovascular disease in London. Barker (2007, p. 412) writes of the intriguing observations they made of the data:

The geography of heart disease in Britain is paradoxical. Rates are twice as high in the poorer areas of the country, and in lower income groups. The steep rise of the disease in Britain and other Western countries was associated with rising prosperity. So why should its rates be lowest in the most prosperous places, such as London and the home counties, and in the highest income groups? …. It is, perhaps, surprising that it was geographical studies that gave the early clue that answers to these paradoxes may come from events in utero.

Looking at these studies, they found that rates of coronary disease were correlated with previous rates, in the same communities, of neonatal mortality, and, significantly, that conditions of adult life did not appear to account for the correlations. For instance, Barker (2007, p. 413) writes, ‘differences in dietary fat consumption did not have the same geographical distribution as past infant mortality’. The correlations could be explained by examining the conditions in which neonatal mortality arose, principally by looking at birth weight and its association with maternal undernutrition. Barker (2007: p. 414) writes:

These geographical studies led to the hypothesis that undernutrition in utero and during infancy permanently changes the body's structure, physiology and metabolism and leads to coronary heart disease and stroke in adult life.

Birth weight, which is associated with a whole host of diseases and chronic conditions in later life including hypertension, diabetes and, indeed, obesity (Phillips, 1998; McMillen et al, 2001; Lau and Rogers, 2004; Salafia et al, 2007; Vikse et al, 2007), is used as a proxy to measure this: babies with low birth weight are understood to have been more nutritionally deprived in the womb than their relatively chubbier counterparts.

Barker's group made numerous studies of the historical data that supported the hypothesis. For example, they uncovered an exception to the epidemiological dogma that predicts correlations between high rates of cardiovascular disease, high rates of neonatal mortality, low birth weight and low socioeconomic status: London's poor slums. The group speculated the explanation must come from common characteristics shared by residents of the slums, which was their rural upbringing. In an article published in The New Yorker, Hall (2007, Paragraph 8) recounts the explanation Barker surmised for the phenomenon:

‘London renewed itself through the nineteenth century by bringing in the fittest young people – men and women, but particularly women who went into domestic service. They came from the villages’, he said, people born and raised in England's prime agricultural country, to the east and south of London. ‘You start from very well-nourished communities, with tall churches and tall people’, Barker said. But because agricultural wages were extremely low, wage riots broke out in the nineteen-twenties, and many young people migrated to London. ‘They brought with them all the advantages of having had a well nourished childhood, and the women gave birth to healthy babies, who tended not to die’, he said.

In other words, although they came to poverty in London and maintained patterns of settlement in its slums, their development occurred in a place that provided adequate nutrition for both themselves and their mothers; such privilege extended to a degree to their own children who, though born into poverty, had the advantage of a mother who was well-nourished during her development in the womb. Much of Barker's work has focused on the effect of maternal nutrition, lifelong and during pregnancy, on foetal development.

This is a population-based understanding of what can be called nutritional privilege, and it illustrates two important innovations of the Barker hypothesis: first, it suggests that responsibility for health and disease is found not only in biological pathways, but is diffused horizontally in the intersections of social, political and economic structures and patterns of human geography that affect maternal nutrition. The concept ‘social determinants of health’ (see Thisted, 2003), a ‘shorthand for describing health approaches that move beyond biomedical and behavioural risk factor approaches to health promotion’ (Raphael, 2006, p. 652), is integral to the Barker hypothesis, which suggests that health, like wealth, is distributed socially according to, for example, gender (Denton et al, 2004; Phillips, 2005; Sen and Ostlen, 2008), race and ethnicity, class (Mackenbach and Howden-Chapman, 2003) and human geography. Although this concept is not new, the move is significant because, although in principle they are concerned with ‘social factors’ in the aetiology of disease, epidemiologists tend to account for them only as statistically confounding variables rather than as causative (Pearce, 1996), especially in the context of what has been called the ‘epidemiology wars’, debates about the best ways to account for multi-causality (see Shim and Thomson, 2010). Pearce (1996, p. 679) argues that most modern epidemiologists consider social aspects only as ‘decontextualised individual risk factors, rather than … population factors in their social and historical context’. Yet according to the hypothesis, these factors are absolutely integral to development and disease.

Second, the Barker hypothesis is novel in suggesting that not only is health distributed socially, but temporally. Certainly time is a significant concept in epidemiology especially in longitudinal population health studies; it is the understanding of the vertical diffusion of responsibility for health through generations that is novel. Both the foetus’ conditions for birth and its mothers’ conditions for birth matter (captured by the title of a lecture delivered by Barker (2010), ‘Your Mother's Mother: The Key to your Health’). Barker (2004, p. 589S) explains:

the ability of a human mother to nourish her baby is partly determined when she herself is in utero, and by her childhood growth, the human foetus is receiving a ‘weather forecast’ based not only on conditions at the time of the pregnancy, but on conditions a number of decades before.

That dependent variables of one epidemiological foray become independent in this way and vice versa reveals them as artifices of methodology.

Social and temporal factors can impact the development of the foetus because of its ‘developmental plasticity’ (Crespi and Denver, 2005) during the critical and largely irreversible period of growth occurring in utero. A plastic organism can develop in response to stimuli that indicate conditions outside the womb, adapting a ‘best fit’ for those conditions (Bateson et al, 2004) such that ‘one genotype can give rise to a range of different physiological or morphological states in response to different environmental conditions during development’ (Barker, 2004, p. 589S). However, importantly for this discussion, developmental plasticity cannot be understood as being strictly advantageous (Barker, 2004). Developmental plasticity is also the capacity of the organism to allot resources selectively and hierarchically (Ellison, 2005). If there is a paucity of nutrients due to factors such as maternal malnutrition, inadequate placentation or some other incidental condition, some organs are favoured over others, the brain in particular (Lumbers et al, 2001). This can lead to lifelong deficits in the functioning of some organs resulting in disease, chronic conditions and early death. There is, for example, a

selective trade-off in the development of tissues that are less important to the baby, such as the kidney. Reduced numbers of nephrons at birth is a life-long deficit, as all nephrons are formed during a sensitive period of development in late gestation. The resulting increased functional demand on each individual nephron, for example by increased blood flow through each nephron, may lead to acceleration of the nephron death that accompanies normal aging. (Barker, 2004, p. 420)

This phenomenon is referred to as ‘foetal programming’, which ‘describes the process whereby a stimulus or insult during a critical period of development has lasting or lifelong effects’ (Godfrey and Barker, 2001, p. 611). In other words, the same plasticity that can afford the ‘best fit’ of the foetus for the environment outside the womb can also be responsible for impaired organ morphology and functioning – a catch-22 strategy for survival Jones characterises as ‘making the best of a bad start’ (Jones, 2005, p. 23).

According to the Barker hypothesis, cases of obesity may be one consequence of developmental plasticity in the milieu of contemporary society (Breier et al, 2001; Fall, 2011). For example, Oken and Gillman (2003) and Kopelman (2000) report on associations between obesity in later life and inadequate development of the pancreas, a key organ regulating digestion and energy storage. Kopelman (2000, p. 639) argues that if there is a paucity of nutrients available to the foetus, it

adapts its growth and metabolism to the expectation of poor availability of nutrition postnatally. This may have survival advantages in utero by targeting available nutrients to essential organs and, in later life, by increasing the ability to store energy as fat to provide energy reserves for use when food is scarce.

This type of foetal programming has been referred to as ‘metabolic imprinting’. The problem, of course, arises when food is not scarce and the adaptations are no longer appropriate. Yet associations between conditions in the womb and obesity are not straightforward. For example, studies have shown an association between low birth weight and impaired development of insulin secretory responses, which could lead to obesity in later adult life; contrariwise, being large-for-gestational age is associated with maternal hyperglycaemia (high blood glucose level) leading to foetal insulin resistance syndrome promoting excess adipose tissue in early and later life (Oken and Gillman, 2003, p. 498).

The two key concepts of the Barker hypothesis – developmental plasticity and foetal programming – provide the basis for alternatives to dominant aetiologies of obesity that surmise ‘human agency’ as the ‘problem’, and do not represent forms of biological essentialism despite what ‘programming’ might suggest. Rather, health and disease are, to use the language of the philosophy of science, emergent: they cannot be attributed to any one part or sum of parts of the system, which are both ‘natural’ and ‘social’ and are not always known, but rather come about because of the interaction of parts at multiple levels (Gilbert and Sarkar, 2000). We can look to the participation of the placenta to deftly illustrate the emergence of development. A mother does not nourish her baby with the food she eats; rather, her metabolism, itself partly ‘programmed’ during her time in utero, determines the way she handles food, and it is the placenta, the means by which the foetus has access to a world beyond itself via the maternal body, that makes these nutrients available. No two placentas function in exactly the same manner, nor does a placenta remain static over time. The placenta is, in many ways, a being unto itself: it can be genetically distinct from the foetus (Kalousek and Vekemans, 1996), ‘pathological’ or become injured. It can compete for resources with the foetus as it, too, requires nutrients to survive and grow. Its development can be hampered by the maternal immune system. Finally, normal or incidental variations in its structurally vulnerable connection to the mother can confer long-term effects on the foetus (Barker, 2004, p. 590S). In other words, the participation of the placenta alone in foetal programming means that health and disease, traceable as they are to the developmental process, are ‘not given in the order of things’ (Law, 1999, p. 3) but complex and emergent. In this way, the hypothesis posits what has been called in science studies a relational materialism of development, health and disease in which every being comes to be ‘through, by, and in’ relations (Law, 1999, p. 4), which are extensively heterogeneous and dynamic.

Obesity is often seen as a symptom of a profoundly pathological culture of excess calories, fat, carbohydrates, laziness and ignorance. Through the lens of the hypothesis, however, the ‘excess’ fat of obesity is perhaps not so much the corporeal manifestation of accumulated poor choices of individuals or their mothers, but rather a materialisation of our social and material situatedness in space and time, indicating diffusive biosocial responsibility that transcends the individual and the present time.

Fat Transdisciplines

The Barker hypothesis therefore has the potential to support FA and critical obesity studies, particularly in framing obesity as an issue of social justice by suggesting that ‘lifestyle’ is not the only or even the best place to intervene as bodies materialise relations, times, places and conditions that are beyond our individual control or grasp. Second, the hypothesis suggests that fatness is not just potentially an issue of health that can be addressed by basic science and clinical medicine alone, but one of social and economic inequity. It sheds light, for example, on how serious nutritional deprivation can be in its health effects through generations. Burke's (2011, p. 222) claim that ‘obesity and food insecurity share many of the same aetiologies’ becomes sensible in such a perspective. Seen in this light, Barker's advocacyFootnote 11 for better nutrition for girls and women is a feminist and social justice-based project indeed. It provides critiques for fat stigma, marginalisation of fat people and uncritical notions of beauty and fitness through refining origins of the accumulation of fat. Finally, although FA is often understood as a position that helps individuals accept themselves, the Barker hypothesis suggests what is needed is a society-wide consciousness of our mutual responsibility for the health of all people, shared with the non-human world that comprises the environment.

For obesity research, this diffusive biosocial responsibility for health and disease means that the possible spaces for action grow along with the aetiological net cast by the Barker hypothesis. Although some of these spaces may be best ascertained by scientific methods and analyses others by social scientific ones, the ways in which the Barker hypothesis figures development suggests that an integrative, transdisciplinary approach that addresses the multifarious conditions of life can better generate explanatory insights. Weasel (2001, p. 28) writes, ‘we can go beyond simply critiquing and take on the task of identifying examples of science that might serve as critical resources for re/constructive feminist [or other critical] projects in the science’, but we can also go a step further and seek integrative crossover that is ultimately mutually transformative of disciplines, an approach that Mackenzie and Murphie (2008; and taken up by Hird, 2009) have referred to as ‘engagement’. Such mutual transformation is crucial to transdisciplinarity, which Nicolescu (2008, p. 2; emphasis added) defines as concerning ‘that which is at once between the disciplines, across the different disciplines, and beyond all disciplines’ to which Songca (2006) adds ‘outside’ disciplines. For Flinterman et al (2001, p. 258), it is an approach that creates ‘integral knowledge’, which ‘is characterised by its stronger orientation toward public perspectives and its problem-solving capability’. In other words, transdisciplinarity exists in the spaces that are not currently occupied by disciplines, that cannot be exhausted by any one discipline alone or that cannot be filled by the mere coexistence, peaceful or not, of disciplines. These spaces, as well as ways forward that can be called truly transdisciplinary, may only become ascertained in the mutual transformation of disciplines in their encounters.

The question becomes whether scientists and social scientists are prepared to undertake this project. On the scientific side, there is considerable scepticism as to whether social scientists can contribute much of value to the project of science. The perception of the social sciences as being ‘soft’ and ‘subjective’ severely impacts the degree to which social scientific knowledge is taken seriously. On the social scientific side, science can be outright ignored, critiqued with asymmetric attention given to the failures of science, or viewed strictly through a social constructionist lens that is unable to account for scientific evidence. These disciplines fail to fully ascertain health and disease because they often take an exclusively ‘natural’ or exclusively ‘social’ world as their object of study. As such, resources in scientific knowledge that may be relevant for social scientific projects are not effectively mined, examined or employed; likewise, science cannot benefit from the contributions of social scientists. Similarly, concurrent projects that address biosocial origins of phenomena of interest, such as those of the behavioural sciences (LaFreniere, 2000; Featherman and Lerner, 1985) or the burgeoning field of epigenetics, cannot be recognised or cross-deployed.

I suggest that avoiding overemphasising the present political impact of the Barker hypothesis or suggesting that it is a panacea for social justice is key. Many important and pointed critiques have been fielded against the hypothesis. For example, Evans (2010) critiques the ‘anticipation of fatness’ inherent in risk assessment as a form of ‘pre-emptive politics’ that supports rather than challenges a notion of fat as pathological. The hypothesis has been criticised for assuming that social structures and one's place within them remain static through the life course and generations (Vagero and Illsley, 1995) and for ‘racialising’ disease in using race as an operative concept (Montoya, 2007; Paradies et al, 2007; Poudrier, 2007). Important critiques address the ways in which the hypothesis is mobilised towards the biomedical control of women's reproductive bodies. Keenan and Stapleton's (2010) study of information about and measurements of pre-pregnancy, pregnancy and post-pregnancy BMI argues that women's bodies are already extensively produced and normalised by disciplinary mechanisms availed in medicine and science. In engaging in the production of knowledge on ‘risk’, the Barker hypothesis becomes moralised as a technology for normative behaviours particularly of mothering. Barker the man has been the object of much criticism, perhaps for a personal tendency to make ‘sweeping statements’ that do not seem in agreement with the hypothesis such as ‘a woman is merely the arena for miracles over which she has no control’ (Hall, 2007, Paragraph 15).

Warin et al's (2011, p. 458) look at the genealogy of the hypothesis shows that it has been absorbed within the behavioural model, as recent DOHaD research has shifted the focus from social determinants to ‘the interior of women's bodies’, framing women as responsible for the ‘transmission’ of obesity through generations (Warin et al, 2012). Likewise, the need for engagement is urgent, as some social scientific knowledge is being swept into the war on obesity. For example, Evans (2010) finds that shifts in focus to obesogenic environments implicates geographers in the politics of obesity, a point inadequately addressed by the discipline. Obesity is already being framed as an issue of environmental justice (see Taylor et al, 2006; Cutts et al, 2009), yet critical geographies must go beyond describing obesogenic environments to critically examining how obesity is experienced in materialising spatiotemporal contexts (Evans and Colls, 2009; Evans, 2010). Evans suggests that a lack of engagement with the temporal dimensions of fat, which as noted is found in the Barker hypothesis, means that a critical politics of risk for obesity is not being developed.

Importantly, it has been charged that some perspectives on fat

substitut[e] ‘bad science’ with ‘good science’ without critiquing the use of science, and much of the discourse relies on an uncritically positivist stance regarding truth and facts … [from which] fat people themselves are abstracted and largely absent from the discourse which often also fails to engage with the broader historical Fat Studies literature and activism. (Cooper, 2010, p. 1023)

In keeping with a transdisciplinary approach in which the mutual interrogation and transformation of disciplines is emphasised, social scientific critiques and approaches must contribute in ways that address these concerns about the significance and deployment of the hypothesis. Only by engaging with the hypothesis in transformative ways can its own transformative potential be actualised.

Conclusion

My argument contains two tendentious points: obesity is a social justice issue, and science, specifically the Barker hypothesis, may support this politicisation of obesity in particular as a resource for both FA and transdisciplinary collaboration in the study of reproduction, disease, lifestyle, life course, generation, nutrition and development. Just as the Barker hypothesis has become critical to research in diverse fields in the sciences, so it should in the social sciences, not only because its approaches are potentially integrative of all sorts of social scientific knowledge – and in fact, sorely in need of them – but also because it is an avenue through which research can be made more responsive and impactful. There are significant barriers to change. Those concerned with social justice may be reluctant to look to the Barker hypothesis for support of their cause, raising questions about the potential for social injustice engendered by the Barker hypothesis. However, I argue that these points are precisely why those working with and through the Barker hypothesis, obesity and fat need transdisciplinary collaboration. Obesity illuminates the limitations of both reductivist scientific and constructivist social scientific approaches. What may emerge from transdisciplinary collaboration is not an integrated ‘theory of everything’ of fat, but rather the recognition of and attendance to the constant biosocial negotiation of corpulence.