Unprocessed Red Meat and Processed Meat Consumption, Plasma Metabolome, and Risk of Ischemic Heart Disease: A Prospective Cohort Study of UK Biobank

Abstract

Background The evidence is equivocal on the association between meat consumption and ischemic heart disease (IHD) risk. To what extent the variation of individuals' metabolic responses to the same diet may account for this association is not fully understood. We aim to identify metabolomic signatures characterizing consumption of unprocessed red meat and processed meat and whether such signatures are associated with IHD risk. Methods and Results We conducted a cohort study of 92 246 individuals (mean age, 56.1 years; 55.1% women) using the UK Biobank. During the median follow-up of 8.74 years, 3059 incident IHD events were documented. Unprocessed red meat and processed meat consumption was assessed using a touchscreen dietary questionnaire. Plasma metabolome was profiled by high-throughput nuclear magnetic resonance spectroscopy. Cox proportional hazards regression model was used to test the association of meat consumption with IHD. Genome-wide association analysis and 1-sample Mendelian randomization were performed for metabolomic signatures and causal association of signatures with IHD. Using elastic net regularized regressions, we constructed metabolomic signatures consisting of 157 and 142 metabolites for unprocessed red meat (Spearman correlation coefficient [r]=0.223) and processed meat (r=0.329), respectively. These signatures showed positive associations with incident IHD (red meat related signature: hazard ratio [HR] per SD increment=1.11 [95% CI, 1.06-1.16], P<0.001; processed meat related signature: HR, 1.16 [95% CI, 1.11-1.21], P<0.001). Genome-wide association studies identified 45 and 4 loci, involved in lipid and lipoprotein metabolism, for red and processed meat related signatures. Mendelian randomization showed that there were casual associations of signatures with risk of incident IHD. Conclusions We identify metabolomic signatures that reflect consumption of unprocessed red meat and processed meat, and these signatures are associated with an increased risk of IHD.

Keywords: Mendelian randomization analysis; genome‐wide association analysis; ischemic heart disease; meat consumption; metabolomics.

Figure 1. Unprocessed red meat and processed meat consumption, plasma metabolome, and risk of ischemic heart disease.

The identified red and processed meat related metabolomic signatures, which mainly consist of lipid metabolites, are associated with an increased risk of ischemic heart disease. In 1‐sample MR analyses, 45 and 4 loci for red meat and processed meat related metabolomic signatures (P<5×10⁻⁸) were used as the instrument variable to predict the metabolomic signatures. Cox proportional hazards model was adjusted for sex, age, race and ethnicity, UK Biobank assessment center, education, body mass index, systolic blood pressure, diastolic blood pressure, smoking status, alcohol intake, employment status, physical activity, the Townsend Deprivation index, fruit and vegetables, family history of cardiovascular disease, baseline diabetes, hyperlipidemia, and cancer. GWAS indicates genome‐wide association study; HR, hazard ratio; IHD, ischemic heart disease; and MR, Mendelian randomization.

Figure 2. Correlation matrix for red and processed meat and all the metabolic biomarkers in the primary analysis.

A, Colors represent directions of the Spearman correlation (blue‐negative and red‐positive), and the color depth represents correlation magnitudes (the darker the stronger). B, The correlations between consumption of red and processed meat and related signatures. FDR indicates false discovery rate; HDL, high‐density lipoprotein; LDL, low‐density lipoprotein; R, Spearman correlation coefficient; and VLDL, very low‐density lipoprotein.

Figure 3. Cumulative risks of ischemic heart disease by red and processed meat and related metabolomic signature groups.

Cox regression models were adjusted for sex, age, race and ethnicity, and UK Biobank assessment center. HR indicates hazard ratio; and IHD, ischemic heart disease.

Figure 4. Distribution and relationship of standardized metabolomic signatures with incident ischemic heart disease risk in the UK Biobank.

Gray bars represent the distribution of metabolomic signature in participants in the UK Biobank. Nonlinear relationship between metabolomic signature and IHD risk was assessed using a restricted cubic spline analysis with knots placed at the 10th, 50th, and 90th percentiles of metabolomic signature, and hazard ratios were estimated with adjustment for sex, age, race and ethnicity, UK Biobank assessment center, education, body mass index, systolic blood pressure, diastolic blood pressure, smoking status, alcohol intake, employment status, physical activity, the Townsend Deprivation index, fruit and vegetables, family history of cardiovascular disease, baseline diabetes, hyperlipidemia, and cancer. HR indicates hazard ratio.