Cardiovascular and metabolic diseases are a primary cause of morbidity and mortality worldwide. Diet is a major risk factor for cardio-metabolic health, but is challenging to study in part because metabolic response to diet is highly individualised. Characterising the molecular pathways that mediate personalised responses to diet is critical to effectively tackle the current epidemic. Epigenetic mechanisms are key regulators of gene function that can change in response to external stimuli, including diet.

The DIMENSION consortium investigated the causal impacts of diet on epigenetic regulation of gene function across tissues, and their impact on subsequent cardio-metabolic health outcomes. The work programme was based on (i) targeted postprandial and nutritional intervention studies to tackle the causal relationships between diet, epigenetic modifications and gene function, in combination with (ii) state-of-the-art analyses to characterize the links between diet, epigenetics, and cardio-metabolic health using novel diet measures including metabolomics in extensively studied cohorts, and (iii) functional follow-up experiments assessing the impact of diet-induced epigenetic and transcriptomic signals at the level of cells, within and across tissues. To this end, the consortium completed epigenetic and transcriptomic profiling from up to three targeted postprandial and nutritional intervention studies and carried out a series of experimental studies in in-vitro systems. To aid large-scale epigenomic data processing and quality assessment DIMENSION consortium partners developed computational tools for methylome data analysis (DNAmArray).

The primary results from DIMENSION consortium identified epigenetic and transcriptomic postprandial molecular trajectories, and identified links between these molecular trajectories and the postprandial glucose response. The DIMENSION project also identified epigenetic changes that occur as a result of a diet and lifestyle intervention in older adults. Subsequent cohort-based analyses further characterised links between diet, epigenetics, and cardio-metabolic health. These included studies identifying epigenetic signatures of diet quality and habitual diet intakes, both at nutrient level and at the level of food groups. Large-scale analyses of cardio-metabolic health identified epigenetic biomarkers of multiple phenotypes, for example, accumulation of visceral fat, circulating lipid levels, and cytokines and adipokines. Follow-up dissection of these effects identified both signals that are putatively causal to cardiometabolic health, those that mediate effects of diet risk factors on metabolic health, as well as signals that appear to be secondary to the phenotypes. Longitudinal epigenetic data profiling in a sample subset identified that a subset of diet-induced epigenetic signals exhibits longitudinal stability. Integrating transcriptomic and metabolomic data allowed for a functional assessment of regulatory genomic diet-induced signals, which was further investigated in a series of in-vitro experiments of diet-induced regulatory genomic effects in cells and tissues.

In summary the DIMENSION consortium aimed to deliver novel mechanistic insights into how exactly nutritional modifications can modulate the regulatory and functional genome to promote cardio-metabolic health. The insights obtained can complement traditional population-based guidelines and have potential impacts to inform the rapidly evolving area of individualised nutrition-based strategies.

Analyses within DIMENSION consortium datasets have contributed towards identifying epigenetic signatures of habitual diet composition and quality (Hellbach et al. 2022, Hellbach et al. 2022, Do et al. 2020 and Ma et al. 2020), many of which were also associated with multiple health outcomes or were only present in individuals who exhibit specific metabolic health characteristics.

The DIMENSION consortium identified an epigenetic signature of visceral fat, a key risk factor for metabolic disease, where some of the signals mediate the effects of diet on metabolic health (Christiansen et al. 2022). Our epigenome-wide analyses of lipid metabolites provide evidence of a link between DNA methylation with lipid compositions and lipid concentrations of different lipoprotein size subclasses, thus offering in-depth insights into well-known associations of DNA methylation with total serum lipids. A larger consortium analysis in currently underway (Gomez Alonso et al. 2020).

Omixer, a Bioconductor package for multivariate and reproducible sample randomization in omics studies has been developed and can be accessed here.