Gut microbiome co-abundance networks varies with age, sex, smoking status and body mass index

The gut microbiome is a complex ecosystem whose members develop local interactions to form coherent functional communities. Variability in these communities, typically investigated through taxa co-abundance, might provide critical insights on the biological links between the gut microbiome and human phenotypes. However, existing methods to investigate variations in taxa co-abundance suffer multiple limitations. Here, we address the simple but challenging question of identifying factors associated with variability in gut microbiome taxa co-abundance using a novel covariance-based method that resolve these limitations. We screened 80 host factors in 938 healthy participants, and identified associations between taxa co-abundance variability and age, sex, smoking status, and body mass index (BMI) not captured by abundance-based and diversity-based methods. Increased age and smoking were associated with an overall decrease in co-abundance, and conversely BMI with an increase. Finally, we demonstrate that the proposed approach offers a powerful framework for describing taxa networks at the individual level and predicting host features.