Association between gut microbiome profiles and host metabolic health across the life course: a population-based study

Abstract

Background: The human gut microbiome changes considerably over time. Previous studies have shown that gut microbiome profiles correlate with multiple metabolic traits. As disease development is likely a lifelong process, evidence gathered at different life stages would help gain a better understanding of this correlation. Therefore, we aim to investigate how the association of the gut microbiome and metabolic traits change over the lifespan. Methods: We identified microbiome patterns (clusters) within two population-based cohorts at different life stages, i.e., pre-adolescents of the Generation R Study (mean age 9.8 years; n = 1488) and older adults of the Rotterdam Study (RS, mean age 62.7 years; n = 1265) using K-Means clustering, and surveyed for host metabolic phenotypes, lifestyles and other factors driving these patterns. Analyses were replicated in the Lifelines-DEEP Study (mean age 45.0 years; n = 1117). The association between microbiome clusters and host metabolic health was evaluated as well as the link between microbiome clusters and incident atherosclerotic cardiovascular disease (ASCVD) in RS during follow-up (median 6.5 years). Findings: We identified two distinct microbiome clusters (U and H) within each study population presenting contrasting metabolic statuses. Cluster U was characterized by lower microbiome diversity, increased Streptococcus, Fusicatenibacter, and decreased Prevotella_9 and Christensenellaceae_R-7_group; wherein individuals showed higher fat percentage, triglycerides, use of medications, and lower socioeconomic status. Individuals in cluster U had increased odds (between 1.10 and 1.65) of being relatively metabolically unhealthy and presented a higher 5-year ASCVD risk (mean risk 0.059 ± 0.071 vs 0.047 ± 0.042, p < 0.001). Interpretation: We provide evidence of a life–course relationship between gut microbiome profiles and metabolic health. Funding: R.L is supported by European Union Horizon 2020 research and innovation program under Marie Skłodowska-Curie grant agreement No 860898 [FIDELIO].