Epigenetic memory or metabolic reprogramming? A new framework sorts the evidence in cardiovascular disease

Not every epigenetic change seen in diseased tissue is a memory. That is the central argument of a review published in Clinical Epigenetics, which sets out to separate two phenomena that are often conflated: persistent metaboloepigenetic states, in which chromatin alterations laid down by past metabolic exposure outlast the exposure itself, and reprogramming under continued stress, in which the epigenome keeps adapting while the damaging stimulus is still present. The authors, Li Lanlan, Guo Fan, Zhang Xiaoqing, Zhao Ran and Fu Jianhua, argue that the distinction matters because previous metabolic exposure can influence cardiovascular risk long after treatment begins, yet the mere presence of chromatin changes during disease does not establish that a memory mechanism is at work.

To make the distinction operational, the review applies eight criteria to 43 study-specific evidence units, scoring persistence, mechanistic causality, cardiovascular relevance and human evidence separately. The framework explicitly distinguishes exposure withdrawal from incomplete recovery, persistent dysfunction from functional recall, and prevention of an epigenetic state’s induction from reversal of a state that has already been established. According to the authors, study-specific timing and intervention data should determine any claim of memory; neither disease association nor an aggregate evidence score can substitute for them.

The strongest human signals come from diabetes cohorts, where histone acetylation and DNA methylation signatures have been linked to previous glycaemia and to subsequent complications. The review is careful about what these findings show: they concern mainly microvascular outcomes and do not, by themselves, establish chromatin-mediated cardiovascular risk. In other words, a molecular fingerprint of past blood sugar levels can be detected in human tissue, and it tracks with later disease, but the causal chain from that chromatin state to a macrovascular event has not been closed.

Complementary evidence of persistence comes from controlled glucose-recovery experiments in human CD34+ progenitors, endothelial cells and a monocytic cell line. These systems allow investigators to withdraw the metabolic stimulus and then ask whether the epigenetic and functional alterations remain, which is precisely the test that observational human cohorts cannot perform. The review treats these controlled studies as providing evidence of persistence that the human association data lack, while noting that they remain cell-based and cannot on their own demonstrate later cardiovascular events in patients.

General trained-immunity models, in which innate immune cells retain a heightened response state after an initial stimulus, serve in the review as a source of mechanistic clarity about the metabolic and chromatin machinery involved. By contrast, models built around oxidised lipoprotein, diabetes and post-infarction settings provide more direct links to the cardiovascular system. This layering matters: a mechanism demonstrated elegantly in a general immune-memory model still needs to be connected to vascular and myocardial biology before it can be called cardiovascular epigenetic memory.

The review also flags two vascular phenomena that it argues should be kept apart. Inflammatory persistence in vascular smooth muscle is one; lactylation-associated calcification is the other. For the latter, the authors state that postwithdrawal memory remains unestablished, meaning that no study has yet shown the calcification-linked chromatin state surviving after the initiating metabolic condition is removed. Lumping the two together, they suggest, risks overstating what the current literature supports.

Mechanical unloading and pressure-relief studies add a further strand of evidence, revealing persistent myocardial features after the heart’s load is reduced. Here too the review urges caution: metabolic causation and chromatin storage of these persistent features are incompletely resolved. A persisting cellular phenotype after unloading is consistent with memory, but without evidence that a chromatin mechanism stores the state and that metabolism caused it, the observation alone does not complete the picture.

Where does this leave the field? The authors conclude that selected immune and vascular experiments do support metaboloepigenetic memory, and that human studies have identified plausible molecular carriers, such as the histone acetylation and DNA methylation signatures seen in diabetes cohorts. What is missing is the final link: a demonstration that a persistent, cell-defined chromatin state causes later cardiovascular events in humans. The assessed evidence, they write, does not complete that causal chain.

The practical payoff of the framework is a research agenda aimed at closing that gap. The authors call for longitudinal, cell-resolved biomarker validation, so that candidate epigenetic marks can be tracked over time in defined cell populations and tested against later outcomes, and for interventions delivered after state establishment, which would directly test whether reversing an established chromatin state changes disease trajectory. Until such studies are done, the review suggests, claims about cardiovascular metabolic memory should be calibrated to what each individual study actually demonstrated, with attention to whether the exposure was withdrawn, whether the state persisted, and whether the mechanism was chromatin-based.

Subject of Research: Distinguishing persistent epigenetic memory from stress-induced reprogramming in cardiovascular disease

Article Title: Distinguishing persistent memory from reprogramming in cardiovascular epigenetics

Article References: Lanlan, L., Guo Fan, Xiaoqing, Z., Ran, Z., & Jianhua, F. (2026). Distinguishing persistent memory from reprogramming in cardiovascular epigenetics. Clinical Epigenetics. https://doi.org/10.1186/s13148-026-02261-3

Image Credits: AI Generated

DOI: 10.1186/s13148-026-02261-3

Keywords: epigenetics, chromatin, cardiovascular disease, metabolic memory, trained immunity, DNA methylation, histone acetylation, diabetes, vascular smooth muscle, myocardial unloading, lactylation, review