Determinants of the plasma metabolome: cross-sectional and longitudinal associations over six years in the NESDA cohort
Summary Background The plasma metabolome represents a valuable molecular readout of a person’s physiological state, yet its relation to health, stress and lifestyle remains underexplored collectively. Methods Here, we conducted an untargeted metabolomics analysis using 3804 paired samples from 1902 participants of the observational Netherlands Study of Depression and Anxiety at baseline and six-year follow-up, quantifying 680 plasma metabolites. We characterised five metabolome principal components, three with distinct biochemical enrichments related to transmembrane transport, sphingolipid, and amino acid metabolism. Findings Metabolite levels showed moderate intrapersonal correlation between baseline and six-year follow-up (ICCmedian = 0.482), and 22% of metabolites showed standardised mean differences >0.2. Multivariate linear modelling on 18 baseline determinants across demographics, psychosocial environment, lifestyle, somatic and mental health explained a maximum of 35% of baseline metabolome PC variance and 12% of six-year change ΔPC variance. Demographic (e.g., sex, age), somatic health (e.g., BMI, medication) and lifestyle factors (e.g., smoking, alcohol intake) demonstrated strong associations both cross-sectionally and longitudinally, while psychosocial factors and mental health contributed minor explained variance in comparison. Interpretation Altogether, our study provides hierarchical insights into the cross-sectional and longitudinal implications of health, stress, and lifestyle exposures for the plasma metabolome. Funding Geestkracht program of the Netherlands Organisation for Health Research and Development, the Dutch Research Council and the Dutch Ministry of Education, Culture and Science, Stress in Action, Amsterdam Neuroscience, ImmunoMIND, the National Institute of Mental Health, National Institute on Ageing and the Foundation for the National Institutes of Health.