Association of Metabolic Diseases and Moderate Fat Intake with Myocardial Infarction Risk

Abstract

Background: Myocardial infarction (MI) can range from mild to severe cardiovascular events and typically develops through complex interactions between genetic and lifestyle factors.

Objectives: We aimed to understand the genetic predisposition associated with MI through genetic correlation, colocalization analysis, and cells' gene expression values to develop more effective prevention and treatment strategies to reduce its burden.

Methods: A polygenic risk score (PRS) was employed to estimate the genetic risk for MI and to analyze the dietary interactions with PRS that affect MI risk in adults over 45 years (n = 58,701). Genetic correlation (rg) between MI and metabolic syndrome-related traits was estimated with linkage disequilibrium score regression. Single-cell RNA sequencing (scRNA-seq) analysis was performed to investigate cellular heterogeneity in MI-associated genes.

Results: Ten significant genetic variants associated with MI risk were related to cardiac, immune, and brain functions. A high PRS was associated with a threefold increase in MI risk (OR: 3.074, 95% CI: 2.354-4.014, p < 0.001). This increased the risk of MI plus obesity, hyperglycemia, dyslipidemia, and hypertension by about twofold after adjusting for MI-related covariates (p < 0.001). The PRS interacted with moderate fat intake (>15 energy percent), alcohol consumption (<30 g/day), and non-smoking, reducing MI risk in participants with a high PRS. MI was negatively correlated with the consumption of olive oil, sesame oil, and perilla oil used for cooking (rg = -0.364). MI risk was associated with storkhead box 1 (STOX1) and vacuolar protein sorting-associated protein 26A (VPS26A) in atrial and ventricular cardiomyocytes and fibroblasts.

Conclusions: This study identified novel genetic variants and gene expression patterns associated with MI risk, influenced by their interaction with fat and alcohol intake, and smoking status. Our findings provide insights for developing personalized prevention and treatment strategies targeting this complex clinical presentation of MI.

Keywords: fat; metabolic syndrome; myocardial infarction; polygenic risk score; precision medicine; single-cell RNA sequencing.

Figure 1

SNP-SNP interactions and PRS and lifestyle interactions. (A) Distribution of SNPs within a 1 Mb window size across chromosomes 1 to 22. SNPs were filtered based on a significance threshold of p-value < 0.0001. Adjacent regions with high SNP density (≥50 SNPs) are highlighted in red, while locations with fewer SNPs are displayed in a gradient from red to blue, indicating decreasing density. SNPs that did not pass the filtering criteria are shown in gray, serving as a reference for genomic regions without significant associations. (B) Interaction of genes: a comprehensive interaction diagram of ten key genetic elements. (C) Frequencies of MI in participants with low, medium, and high polygenic risk score (PRS) based on the optimal 9 risk alleles: LNX1_rs2616417, ELOVL2_rs75105616, SGCZ_rs73201298, KIFBP_rs3864814, MKRN3_rs56730421, CHD2_rs201915192, RNF213_rs1410411669, RPTOR_rs7224758, and DDC_rs77235945. (a) Participants categorized by alcohol intake (cut-off: 30 g/d). (b) Participants categorized by smoking status (non-smokers versus current and former smokers). (c) Participants categorized by fat intake (cut-off: 15% of daily energy consumption). Significant differences between low- and high-PRS groups: * p < 0.01, ** p < 0.001. MI: myocardial infarction; LNX1: ligand of numb-protein X 1; ELOVL2: ELOVL fatty acid elongase 2; SGCZ: sarcoglycan zeta; KIFBP: KIF-binding protein; MKRN3: makorin ring finger protein 3; CHD2: chromodomain-helicase-DNA-binding protein; CBX2: chromobox protein homolog 2; RNF213: ring finger protein 213; RPTOR: regulatory-associated protein of mTOR; DDC: dopa decarboxylase.

Figure 2

Genetic correlations in Asian ancestry between MI (n = 28,030) and other traits after Bonferroni correction. The traits associated with MI are divided into three categories: dark blue, brown, and green color bars indicate health indicators, dietary factors, and lifestyle factors, respectively. MI: myocardial infarction.

Figure 3

Colocalization analysis of SNPs associated with MI in eQTL datasets. (A) Schematic diagram of colocalization analysis under different hypotheses (H0, H1/H2, H3, H4). A binary vector representing the number of shared variants in each feature’s region. The value on the y-axis indicates whether the variation has a causal relationship with the disease. Matching positions of eQTL (red) and biomarkers (blue) indicate the same causal SNP, while different positions indicate that the causal SNP of the dataset is different. (B) Combined Manhattan plot showing 12 mapped distinct loci and 28 genome-wide significantly associated SNPs (p < 5 × 10⁻⁸), as well as the number of suggestive SNPs identified in each genome. (C) Data supporting a single variant (PP4 > 80%) affecting both traits are identified by a red border. High association evidence genes (eQTL ± 1MB) from extended range analysis are marked as moderate (*) or strong (**). −Log ₁₀ FDR p plots of eQTL representing the tissue expression of significantly associated SNPs with corresponding posterior probabilities for GTEx. (D) LocusZoom plots mapping the genomic locations of significantly associated SNPs (rs3864814, rs2081208) on chromosome 10 and chr16, providing reliable evidence supporting a colocalization signal on STOX1-VPS26A and RP11-744D4.2. (E) eQTL association plots of colocalization of rs3864814, rs2081208 with STOX1, RP11-744D4.2 in the corresponding GTEx dataset. GWAS −log ₁₀ p for SNPs corresponding to GTEx, eQTL −log ₁₀ p for STOX1 and RP11-744D4.2, and evidence of causal variation in posterior probabilities are shown, respectively. SNPs: single-nucleotide polymorphisms; MI: myocardial infarction; eQTL: expression quantitative trait loci; PP4: posterior probability 4; FDR: false discovery rates; STOX1-VPS26A: storkhead box 1_vacuolar protein sorting-associated protein 26A; GTex: Genotype-Tissue Expression.

Figure 4

UMAP and gene expression analysis of cardiac cell types and disease relevance scores for MI. (A) UMAP visualization of major cardiac cell types; UMAP plot showing 12 major cardiac cell types identified from the Heart Cell Atlas v2 dataset. Each color represents a different cell type, including cardiomyocytes, fibroblasts, endothelial cells, mural cells, and others. (B) UMAP of disease relevance scores across cardiac cell types. Cells with higher scores are marked by warmer colors, indicating a stronger association with disease relevance. (C) Violin plot of disease relevance scores by cell type. The width of each violin corresponds to the frequency of cells with scores in each cell type. (D) Dot plot of key gene expression across cardiac cell types. The size of each dot represents the expression level, while the color intensity indicates the relative gene expression in each cell type.