Associations of coffee, alcohol, medication and supplement use with the metabolome and lipidome: an observational study of premenopausal women

Abstract Introduction Lifestyle factors have been consistently linked to various health outcomes. However, few studies have comprehensively assessed how multiple of these factors are associated with metabolome and lipidome in a single study. Objectives We leveraged untargeted metabolomics and lipidomics to determine the associations of frequencies of alcohol and coffee intake, supplement and medication use, with metabolites and lipid species. Methods This study consists of 702 premenopausal women who provided fasting blood samples. Final analysis included 857 metabolites and 828 lipid species. Multivariable linear regression models adjusted for confounders were used to investigate the associations. We corrected for multiple testing using Benjamini–Hochberg false discovery rate (FDR adjusted p-value < 0.01) and Bonferroni correction (Bonferroni-adjusted p-value < 10 − 5 ). Results At FDR p-value < 0.01, the frequency of alcohol intake was associated with the largest number of biomarkers, including 89 lipid species (9 after Bonferroni correction) across several pathways (ceramides, phospatidylcholines, triacylglycerols, and androgenic steroids) and 123 (13 after Bonferroni correction) metabolites. Frequency of coffee intake was associated with 46 metabolites (27 after Bonferroni correction), with the strongest associations observed for quinate, 3-hydroxypyridine sulfate and trigonelline (N′-methylnicotinate). Frequency of multivitamin use was associated with 18 metabolites (6 after Bonferroni correction). Frequency of statin use was associated with 35 lipid species, respectively, but these were not significant after Bonferroni correction. Conclusions Alcohol, coffee, and multivitamins are associated with several metabolites and lipid species. These findings offer valuable insights into potential biomarkers of exposure and molecular pathways through which these exposures may be associated with health outcomes.