Assessing diet-microbiome associations with linear models: variability across human studies

Understanding and identifying relationships between dietary intake and the gastrointestinal microbiome can provide critical information for optimization of human health. However, methodological approaches largely differ and can impact research findings. Here we outline differences in reported diet-microbiome associations from a range of studies utilizing comparable statistical methods. From a subset of 25 studies using comparable methodology, only 3% of reported diet-microbe associations were common across more than one study. As such, we emphasize the inconsistencies of findings within the literature and wider implications of these differences.