Causal impact of genetically-determined fish and fish oil intake on epigenetic age acceleration and related serum markers.

Background: The interplay between diet and healthspan is a topic of great interest in biomedical research. Toward this end, consumption of marine omega-3 fatty acids is of particular significance, as reports suggest that diets focused on seafood can prolong the disease-free portion of the human lifespan. Fish consumption has also been linked to reduced biological aging as measured by epigenetic clocks, but there is no conclusive evidence of a causal relationship. Moreover, fish oils reduce triglycerides, and may affect other lipid profiles, as well as systemic inflammation. To investigate further, we used two-sample Mendelian randomization to investigate potential causality between fish intake and healthspan markers.

Methods: Bidirectional Mendelian randomization was performed in the two-sample setting with publicly available GWAS summary statistics. GWAS data from the UK Biobank for oily fish consumption (n = 460,443) and fish oil supplementation (n = 461,384) were used as the primary exposures. First-generation epigenetic clocks Hannum age and intrinsic epigenetic age acceleration (IEAA), as well as second-generation clocks GrimAge and PhenoAge were collected from an independent dataset of individuals of European ancestry (n = [34,449-34,667]). Finally, data from the Integrative Epidemiology Unit database was used for serum proxies of lipidemia and systemic inflammation (n = [61,308-78,700]). Additional sensitivity analyses, such as reverse causation testing and the Cochran's Q test were performed for exposure-outcome pairs where the inverse variance weighted (IVW) method was significant (p-value < 0.05), and where the MR Egger method indicated an effect in the same direction as the IVW result.

Results: We report that oily fish consumption appears to decrease PhenoAge acceleration (p < 0.0086), whereas fish oil supplementation appears to decrease GrimAge (p $=$ 0.037). Both omega-3 exposures modify the epigenetic clocks in the expected negative, or age-decelerating, direction. For the serum biomarkers, we find evidence that fish oil consumption leads to a reduction in triglycerides (p $=$ 0.004), although HDL and LDL were not significantly modified. Finally, we also detected a suggestive inverse relationship between oily fish consumption and hsCRP (p $=$ 0.064).

Conclusions: Our analysis shows that consuming fish oil, whether through whole food or as a supplement, can have a rejuvenating impact as measured by PhenoAge and GrimAge acceleration. We have also provided evidence further linking fish oil intake and lower triglyceride levels. These results, based on robust MR-based analyses, emphasize the effectiveness of dietary choices in modifying emerging measures of healthspan.