Abstract TU103: Adherence to the Mediterranean diet and some related serum metabolites are strongly associated with epigenetic biomarkers of aging in Mediterranean subjects

Background and aims: The Mediterranean diet (MedDiet) is a healthy dietary pattern that has been consistently associated with reduced cardiovascular disease incidence. Likewise, higher adherence to the MedDiet has been related to beneficial effects on cardiometabolic health, reducing serum inflammatory and oxidative stress biomarkers (i.e GlycA and branched chain amino acids-BCAAs). In addition, a higher adherence to the MedDiet has been associated with higher serum polyunsaturated fatty acids (PUFAs), mainly Omega-3. However, few studies in Mediterranean populations have investigated these associations with epigenetic biomarkers of aging. Our objective was to analyze the association between MedDiet adherence and several generations of epigenetic biomarkers of aging and to examine the relationship between serum GlycA and BBCAs and the aging biomarkers in Mediterranean subjects. Methods: We analyzed participants from two Mediterranean Spanish populations recruited in Valencia. Cohort 1 consisted of high cardiovascular risk subjects aged 55-75 y (n=426), and Cohort 2 (n=200) consisted of healthy subjects from the general population (18-80y). Likewise, DNA was isolated from blood, and DNA methylation was analyzed with the EPICv1 array. After quality control and normalization, several generations of biomarkers of aging were computed using the UCLA and Biolearn platforms for both populations. We focused on Horvath, PhenoAge, GrimAge, DunedinPACE, and CausAge clocks. Adherence to the MedDiet was measured with the validated 17-item MEDAS score (only in cohort 1). Multivariate adjusted regression models were computed. Results and conclusions: In the high cardiovascular risk cohort, the adherence to MedDiet was inversely associated with all the biomarkers of aging examined (Horvath, PhenoAge, GrimAge1, GrimAge2, DunedinPACE, and CausAge). In the adjusted model (with adherence to MeDiet as continuous variable), the most significant associations were for GrimAge1 (P<0.001), GrimAge2 (P=0.001), CausAge (P=0.002), and DunedinPACE (P=0.019). Likewise, in this high-cardiovascular risk cohort we found strong inverse associations of serum PUFA, mainly % of Omega-3 with biomarkers of aging (even after adjustment for MedDiet adherence). In the general population we also observed the strong inverse association between PUFAs and the biomarkers of aging (all P<0.05). Direct associations between GlycA and serum BBCAs with epigenetic biomarkers of aging were detected (P<0.05).