Epigenome-wide association study of dietary fatty acid intake

Abstract

Background: Dietary intake of n-3 polyunsaturated fatty acids (PUFA) may have a protective effect on the development of cardiovascular diseases, diabetes, depression and cancer, while a high intake of n-6 PUFA was often reported to be associated with inflammation-related traits. The effect of PUFAs on health outcomes might be mediated by DNA methylation (DNAm). The aim of our study is to identify the impact of PUFA intake on DNAm in the Cooperative Health Research in the Region of Augsburg (KORA) FF4 cohort and the Leiden Longevity Study (LLS).

Results: DNA methylation levels were measured in whole blood from the population-based KORA FF4 study (N = 1354) and LLS (N = 448), using the Illumina MethylationEPIC BeadChip and Illumina HumanMethylation450 array, respectively. We assessed associations between DNAm and intake of eight and four PUFAs in KORA and LLS, respectively. Where possible, results were meta-analyzed. Below the Bonferroni correction threshold (p < 7.17 × 10^-8), we identified two differentially methylated positions (DMPs) associated with PUFA intake in the KORA study. The DMP cg19937480, annotated to gene PRDX1, was positively associated with docosahexaenoic acid (DHA) in model 1 (beta: 2.00 × 10^-5, 95%CI: 1.28 × 10^-5-2.73 × 10^-5, P value: 6.98 × 10^-8), while cg05041783, annotated to gene MARK2, was positively associated with docosapentaenoic acid (DPA) in our fully adjusted model (beta: 9.80 × 10^-5, 95%CI: 6.25 × 10^-5-1.33 × 10^-4, P value: 6.75 × 10^-8). In the meta-analysis, we identified the CpG site (cg15951061), annotated to gene CDCA7L below Bonferroni correction (1.23 × 10^-7) associated with eicosapentaenoic acid (EPA) intake in model 1 (beta: 2.00 × 10^-5, 95% CI: 1.27 × 10^-5-2.73 × 10^-5, P value = 5.99 × 10^-8) and we confirmed the association of cg19937480 with DHA in both models 1 and 2 (beta: 2.07 × 10^-5, 95% CI: 1.31 × 10^-5-2.83 × 10^-5, P value = 1.00 × 10^-7 and beta: 2.19 × 10^-5, 95% CI: 1.41 × 10^-5-2.97 × 10^-5, P value = 5.91 × 10^-8 respectively).

Conclusions: Our study identified three CpG sites associated with PUFA intake. The mechanisms of these sites remain largely unexplored, highlighting the novelty of our findings. Further research is essential to understand the links between CpG site methylation and PUFA outcomes.

Keywords: DNA methylation; Docosapentaenoic acid; EWAS; Eicosadienoic acid; Eicosapentaenoic acid; Fatty acids; PUFA n-3; PUFA n-6; Stearidonic acid.

Fig. 1

EWAS Workflow: assessing PUFA Intake's Impact on DNA Methylation Using KORA FF4 and LLS Cohorts. KORA: Cooperative Health Research in the Region of Augsburg, PUFA: Polyunsaturated fatty acid, ALA: Alpha-linolenic acid, SDA: Stearidonic acid, EPA: Eicosapentaenoic acid, DPA: Docosapentaenoic acid, DHA: Docosahexaenoic acid, LA: Linoleic acid, EDA: Eicosadienoic acid, ARA: Arachidonic acid, LLS: Leiden Longevity Study

Fig. 2

Manhattan plot of meta-EWAS results of EPA Panel A. Manhattan plot of results from meta-analysis of epigenome-wide association studies on EPA using linear regression models adjusting for age, sex, BMI, smoking, WBC% and technical variables. The x-axis shows the chromosomal position, and the y-axis the -log10 p value of the DMP-PUFA association. The horizontal gray line indicates the genome-wide significance threshold at a Bonferroni-corrected p value lower than 0.05 (alpha = 1.23 × 10 − 7). The red dot represents the significant DMP identified in this analysis labeled with the cpg name and its annotated gene name. Panel B. Manhattan plot of results from meta-analysis of epigenome-wide association studies on EPA using linear regression models adjusted for age, sex, BMI, smoking, WBC%, technical variables, physical activity, energy intake, estrogen therapy and PUFA supplement intake