Metagenomic Analysis Reveals the Distribution of Antibiotic Resistance Genes in a Large-Scale Population of Healthy Individuals and Patients With Varied Diseases

The human gut microbiome is a reservoir for antibiotic resistance gene (ARG). Therefore, characterizing resistome distribution and potential disease markers can help manage antibiotics at the clinical level. While much population-level research has highlighted the strong effect of donor geographic origin on ARG prevalence in the human gut, little is known regarding the effects of other properties, such as age, sex, and disease. Here we employed 2,037 fecal metagenomes from 12 countries. By quantifying the known resistance genes for 24 types of antibiotics in each community, we showed that tetracycline, aminoglycoside, beta-lactam, macrolide-lincosamide-streptogramin (MLS), and vancomycin resistance genes were the dominant ARG types in the human gut. We then compared the ARG profiles of 1427 healthy individuals from the 2,037 samples and observed significant differences across countries. This was consistent with expectations that regional antibiotic usage and exposure in medical and food production contexts affect distribution. Although no specific uniform pattern of ARG was observed, a significant increase in resistance potential among multiple disease groups implied that the disease condition may be another source of ARG variance. In particular, the co-occurrence pattern of some enriched bacterial species and ARGs that were obtained in type 2 diabetes (T2D) and liver cirrhosis patients implied that some disease-associated species may be potential hosts of enriched ARGs, which could be potential biomarkers for the prediction and intervention of such diseases. Overall, our study identifies factors associated with the human gut resistome, including substantial effects of region and heterogeneous effects of disease status, and highlights the value of ARG analysis in disease research and clinical applications.