Using meta‐analysis to understand the impacts of dietary protein and fat content on the composition of fecal microbiota of domestic dogs (Canis lupus familiaris): A pilot study

Abstract

AbstractThe interplay between diet and fecal microbiota composition is garnering increased interest across various host species, including domestic dogs. While the influence of dietary macronutrients and their associated microbial communities have been extensively reviewed, these reviews are descriptive and do not account for differences in microbial community analysis, nor do they standardize macronutrient content across studies. To address this, a meta‐analysis was performed to assess the impact of dietary crude protein (“protein”) and dietary crude fat (“fat”) on the fecal microbiota composition in healthy dogs. Sixteen publications met the eligibility criteria for the meta‐analysis, yielding a final data set of 314 dogs. Diets were classed as low, moderate, high, or supra in terms of protein or fat content. Sequence data from each publication were retrieved from public databases and reanalyzed using consistent bioinformatic pipelines. Analysis of community diversity indices and unsupervised clustering of the data with principal coordinate analysis revealed a small effect size and complete overlap between protein and fat levels at the overall community level. Supervised clustering through random forest analysis and partial least squares‐discriminant analysis indicated alterations in the fecal microbiota composition at a more individual taxonomic level, corresponding to the levels of protein or fat. The Prevotellaceae Ga6A1 group and Enterococcus were associated with increasing levels of protein, while Allobaculum and Clostridium sensu stricto 13 were associated with increasing levels of fat. Interestingly, the random forest analyses revealed that Sharpea, despite its low relative abundance in the dog's fecal microbiome, was primarily responsible for the separation of the microbiome for both protein and fat. Future research should focus on validating and understanding the functional roles of these relatively low‐abundant genera.