In fact, the development of epigenetic biomarkers of disease often pulls in a different direction. Specifically, one that departs from a complex and socially embedded view of biology. Let us again take the example of cardiovascular conditions. No doubt, epigenetic information can have an immediate impact on disease management. Epigenetic biomarkers are already at the forefront of clinical applications in precision cardiology [15]. These markers promise to contribute to patient diagnosis, prognosis, theragnosis, and therapy in several cardiovascular conditions—such as coronary artery disease, hypertrophic cardiomyopathy, acute myocardial infarction, and heart failure (Reviewed in [16, 17]). Although the additional benefits of using these biomarkers are still debated, epigenetic modifications find an increasing association with cardiovascular diseases and contribute to the quest for their early detection, risk prediction, and prevention [18, 19]. Of note, cardio-epigenetic biomarkers often do so by pointing to the mixed genetic and environmental etiology of these conditions: For instance, the combination of genetic and epigenetic information using machine learning has been shown to improve risk score classifiers and predictors for coronary heart disease [13]. But, what about the role of this kind of research in producing a full appreciation of the whole biological-and-social spectrum of factors shaping individual health?
Several social scientists have criticized how epigenetic research treats the social, environmental, and temporal modulators of disease (risk and etiology). Among these critical scholars, there is a fear of novel forms of reductionism in epigenetics that are no less worrying than those attributed to genetics. What if, critics ask, more than non-gene-centric biology, epigenetics turned out to be the science of a miniaturized and molecular version of the environment [20, 21]? Its insights and implications for an environmentally embedded view of health cannot be fully grasped if this science reduces biosocial loops to either methylation risk scores alone, or to “the effects [on human bodies] of proximate variables” [21, p. 17] such as socioeconomic status [10] or scores of traumatic stress [22]. We subscribe to a different (yet related) version of this criticism, which does not consider the reductionism of the life sciences problematic per se. This line of criticism rather asks whether the repertoire of tools and interventions of epigenetics can be expanded to complexify its grasp of biosocial processes of health differentiation. We acknowledge that available lab methods, statistical tools, populational measures, epidemiological questionnaires, and randomized clinical trials may not be suitable to fully dissect these biosocial loops. Furthermore, we are also aware of the clinical utility of ready-made, validated risk classifiers and predictors for the management of health conditions (such as cardiovascular diseases) [13]. Yet, we ask: are the existing methods and approaches in the field fated to overlook the thick social and cultural underpinnings of these conditions [23]? Is an approach that combines the scalability and predictive value of molecular tests with thicker stratifications through social markers amenable to experimentation in epigenetics? There might be a complexity in biosocial views of health that is incommensurable to the ordinary tools of the biomedical sciences. Yet, this does not mean the results of epigenetics ought necessarily to provide a poor picture of the environmental embeddedness of our health. Especially because the imbrication of these biological, psychological, social, and political factors represents an untapped potential for early prevention, risk prediction, and intervention [13].
While being general critiques of epigenetics, these concerns can be extended to cardiovascular epigenetics too. Cardiovascular epigenetic biomarkers certainly show promising healthcare applications [24], yet one could point to several elements that call for improved consideration of a biosocial perspective in their development. First, epigenetic studies of cardiovascular phenotypes are skewed in favor of an understanding of biomarkers as mere targets for molecular and, specifically, pharmacological intervention (epi-drugs). A growing literature points, in fact, to the pharmacological actionability of epigenetic differences, as in the reported case of the drug-tailoring based on genome-wide DNA methylation differences in hypertension [25]. A simple Web of Science searchFootnote 2 shows that therapeutic or pharmacological approaches to epigenomic signatures of cardiovascular diseases get considerably more attention than the actionability of these markers in primordial prevention. Plus, such preventive measures seem even more neglected when we consider that only a minority of the publications mentioning “prevention” does so without also including therapeutics as the potential application of epigenetic knowledge.Footnote 3 While these indicative bibliometric measures cannot account for the nuances of a scientific debate, they suggest that cardio-epigenetic research emphasizes far less a socially embedded view of these risk factors than it focuses on pharmacological interventions to correct them.
Second, the understanding of the environment in this literature displays several limitations in light of a biosocial perspective on cardiovascular diseases. Some studies operationalize the environment just as light, temperature, and food [26]; others through proxy measures of social conditions such as socioeconomic status [10]. Some do underline the multiple pathways and loops between social interactions, psychological stress, and epigenetic predispositions to cardiovascular disease. Yet, they often do so only in animal experiments [27] or as part of conceptual discussions with little practical implementation [28, 29]. The integration of finer-grained measures of the environment and social conditions has found little translation in experimentation. Few studies exist that: (i) explore epigenetic mechanisms by which social influences (e.g., racialized inequalities in the USA) can become embodied health predispositions (e.g., racial disparity in cardiovascular risk and disease) [30]; (ii) probe the cumulative effect of social conditions, inequalities and exposures (e.g., pollutants, chemical hazards) in the (epigenetic) patterning of cardiovascular diseases in our societies [31]; (iii) explore the distinct associations and combinations of biological and proximal (e.g., lifestyle) or distal (e.g., social structures, environmental exposures) risk factors [13]. The lack of integration of these complex views of social–biological transitions in epigenome-wide association studies (EWAS) is a methodological gap that has found recognition only recently—including on this journal [10, 12]. Little consideration is given also, in the EWAS literature, to the need of differentiating the degrees of specificity, stability, and reversibility of epigenetic modifications (e.g., DNA methylation differences) in the face of clinical, behavioral, or social interventions [10, 11]. For instance, few studies have tried to dissect the age-specific associations between DNA methylation differences and cardiovascular phenotypes: The few results available suggest that epigenetic differences may be less relevant to predict cardiovascular outcomes in children than they are in adults [32]. In a nutshell, empirical research on social–biological loops producing epigenetic predispositions to cardiovascular diseases is limited. And, notably, this is due to the lack of fine-grained measures of the multiple sources, effects, temporalities, and mechanisms of the social exposures that produce these biological differences.