The link between the gut microbiota and type 2 diabetes (T2D) warrants further investigation because of known confounding effects from antidiabetic treatment. Here we profiled the gut microbiome in a discovery (n=1900) and validation (n=800) cohort comprising Swedish subjects naive for diabetes treatment and grouped by glycemic status. We will investigate shifts in the microbiome and metabolome of groups with impaired glucose tolerance, combined glucose intolerance, or T2D compared with groups with normal glucose tolerance. Multivariate analyses and machine-learning microbiome/metabolome models will be evaluated as contributors to the microbial variation. Therefore, our study will be able to indicate if the microbiome and/or metabolome can be use as prognostic markers for development of type 2 diabetes, but also proved mechanistic insights to how microbiome and the resulting metabolites may contribute to disease development.