The dispute began with Chiapperino and Paneni’s earlier paper, Why epigenetics is (not) a biosocial science and why that matters, which examined whether and how epigenetic research can genuinely capture the entanglement of biological and social processes in health and disease. Dupras responded with a critical assessment, reading their argument as a call for less reductionism and more holistic methods in the field. That reading, the two authors now insist, misrepresents their intent. They never claimed that epigenetics should be assimilated to nurture while genetics stands for nature, nor did they seek to draw a sharp line between biosocial epigenetics and a supposedly non-biosocial genetics.

At the heart of the reply is a reframing of the central question. The authors argue that the problem is not whether the processes of development, health differentiation, and evolution are either natural or social, either biosocial or not. Instead, they ask to what extent the methods, tools, and study designs of epigenetics can capture the interactions that necessarily occur between nature and nurture, between biological and social processes. This developmental version of epigenetics, which they align with work by researchers such as Robert Lickliter and David Witherington, is explicitly designed to move beyond the inherited dichotomies of the nature-nurture debate rather than reproduce them under a new molecular vocabulary.

The authors also reject the charge that they are demanding a wholesale turn to holism. They acknowledge that the full complexity of biosocial processes may be incommensurable with the methods of the biomedical sciences, yet they explicitly disavow any foundational endeavor toward an all-encompassing holistic biosocial science. Drawing on the history, philosophy, and social anthropology of science, including the work of Hans-Jörg Rheinberger, John Law, and Annemarie Mol, they contend that the reductionism-holism dichotomy is itself inadequate for understanding postgenomic scientific practice. The point, they write, is not whether epigenetic methods are reductionist, since they plainly are, but how they operate most prolifically at the fuzzy boundary between the trivial and the complex.

To make this abstract argument concrete, Chiapperino and Paneni turn to the measurement tools that dominate epigenetic studies of psychosocial environments: stress scales and allostatic load indices. These instruments, widely used to link adversity to molecular change, have drawn sustained criticism for lacking definition, standardization, and fidelity, and for what anthropologist Jörg Niewöhner has called the molecularization of biography and milieu. The authors concede that such methods are certainly reductionist. But they ask whether reductionism necessarily dooms these tools to a poor picture of the environmental embeddedness of health, and their answer is a qualified no.

The problem, they argue, lies less in the instruments than in the study designs that house them. Much epigenetic research reproduces a simplistic linearity: exposures, objectified through stress or allostatic load measurements, produce biological differences readable in the epigenome, which are in turn linked to disease through mechanistic or statistical associations. A different configuration employing the very same reductionist tools could avoid this pitfall. Repeated longitudinal measures of psychological and physiological scales, for example, could reveal the multidirectional effects of stress, its biological embodiment over time, and the biopsychosocial looping effects through which people respond to and reshape their own conditions. Such designs would leave the reductionist qualities of the methods untouched while dramatically complexifying the account of the phenomenon they can offer.

The reply also addresses Dupras’s defense of ambitious large-scale initiatives such as the International Human Epigenome Consortium. Chiapperino and Paneni are careful to state that they do not question the consortium’s importance for advancing standards, methods, reference knowledge, and a sense of global community among epigeneticists. Nor do they deny that the boundary between epigenetics and genetics is blurred, a definitional controversy that continues to occupy commentators such as Bernhard Horsthemke. Yet they maintain that a theoretically modest point remains highly relevant for any endeavor seeking to build reference epigenomes for the understanding of health and disease.

That point comes into sharp focus around ethnicity and DNA methylation. Few researchers hold that interindividual differences in the epigenome can be fully recapitulated by genetic variation, including variation across so-called genetic ancestry groups. The authors cite a study by James Galanter and colleagues showing that, across the genome, genetic ancestry accounts for roughly three-quarters of the association between ethnicity and methylation differences, with the remaining quarter attributed to shared environmental, social, and cultural factors within groups. In that residual quarter lies evidence that both ancestry and the social construction of ethnic difference shape epigenetically measurable disparities in disease prevalence and health trajectories. This finding, they argue, validates ongoing calls to diversify reference epigenomic maps and epigenome-wide association studies beyond their current overrepresentation of participants of European ancestry, but it also demands more than genetic imputation alone.

Here the authors pose a pointed question to the global epigenetics community: what is currently on the agenda to account for the share of health disparities across social groups that stems from differences in social and environmental exposures rather than genetic ancestry? They suggest that the biosocial dimensions of ethnic differences in the epigenome likely lack a fixed ontology, being neither purely biological nor purely social. For some traits, reference epigenomes may prove highly sensitive to genetic differences across ancestry groups; for others, they may be more open to varied developmental trajectories, local contexts, and culturally situated practices. They caution, following Maurizio Meloni and colleagues, that even simplistic cause-effect models can constitute a problematic form of biosocial determinism around epigenetic differences.

Ultimately, Chiapperino and Paneni frame their reply as a call to symmetry: a demand that the development of epigenetic tools, methods, and study designs refrain from overlooking either side of the biosocial continuum. They do not claim to offer a unique path, nor to certify whether epigenetics is or is not a holistic biosocial science. The simplistic stance, they suggest, belongs to those who feel compelled to pick a side in what is ultimately a fabricated dichotomy between the biological and social origins of epigenetic differences. In a field increasingly asked to explain how poverty, discrimination, and stress become embodied, that insistence on balance may prove more consequential than any single methodological prescription, reminding researchers that the most rigorous science of the epigenome may be the one that refuses to let nature and nurture compete for exclusive ownership of it.

Subject of Research: The role of reductionist and holistic methods in capturing biosocial interactions in epigenetic research

Article Title: In medio stat virtus? A reply to Dupras

Article References: Chiapperino, L., & Paneni, F. (2023). In medio stat virtus? A reply to Dupras. Epigenetics Communications, 3(1), Article 5. https://doi.org/10.1186/s43682-023-00017-1

Image Credits: AI Generated

DOI: 10.1186/s43682-023-00017-1

Keywords: epigenetics, biosocial science, reductionism, holism, DNA methylation, allostatic load, genetic ancestry, health disparities, International Human Epigenome Consortium, nature and nurture, study design, philosophy of science