Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Comparative Study
, 13 (8), e0201279

The Fecal Microbiome and Metabolome Differs Between Dogs Fed Bones and Raw Food (BARF) Diets and Dogs Fed Commercial Diets

Comparative Study

The Fecal Microbiome and Metabolome Differs Between Dogs Fed Bones and Raw Food (BARF) Diets and Dogs Fed Commercial Diets

Milena Schmidt et al. PLoS One.


Introduction: Feeding a Bones and Raw Food (BARF) diet has become an increasing trend in canine nutrition. Bones and Raw Food diets contain a high amount of animal components like meat, offal, and raw meaty bones, combined with comparatively small amounts of plant ingredients like vegetables and fruits as well as different sorts of oil and supplements. While many studies have focused on transmission of pathogens via contaminated meat and on nutritional imbalances, only few studies have evaluated the effect of BARF diets on the fecal microbiome and metabolome. The aim of the study was to investigate differences in the fecal microbiome and the metabolome between dogs on a BARF diet and dogs on a commercial diet (canned and dry dog food).

Methods: Naturally passed fecal samples were obtained from 27 BARF and 19 commercially fed dogs. Differences in crude protein, fat, fiber, and NFE (Nitrogen-Free Extract) between diets were calculated with a scientific nutrient database. The fecal microbiota was analyzed by 16S rRNA gene sequencing and quantitative PCR assays. The fecal metabolome was analyzed in 10 BARF and 9 commercially fed dogs via untargeted metabolomics approach.

Results: Dogs in the BARF group were fed a significantly higher amount of protein and fat and significantly lower amount of NFE and fiber. There was no significant difference in alpha-diversity measures between diet groups. Analysis of similarity (ANOSIM) revealed a significant difference in beta-diversity (p < 0.01) between both groups. Linear discriminant analysis effect size (LefSe) showed a higher abundance of Lactobacillales, Enterobacteriaceae, Fusobacterium and, Clostridium in the BARF group while conventionally fed dogs had a higher abundance of Clostridiaceae, Erysipelotrichaceae, Ruminococcaceae, and Lachnospiraceae. The qPCR assays revealed significantly higher abundance of Escherichia coli (E. coli) and Clostridium (C.). perfringens and an increased Dysbiosis Index in the BARF group. Principal component analysis (PCA) plots of metabolomics data showed clustering between diet groups. Random forest analysis showed differences in the abundance of various components, including increased 4-hydroxybutryric acid (GBH) and 4-aminobutyric acid (GABA) in the BARF group. Based on univariate statistics, several metabolites were significantly different between diet groups, but lost significance after adjusting for multiple comparison. No differences were found in fecal bile acid concentrations, but the BARF group had a higher fecal concentration of cholesterol in their feces compared to conventionally fed dogs.

Conclusion: Microbial communities and metabolome vary significantly between BARF and commercially fed dogs.

Conflict of interest statement

Concerning possible competing interests, we would like to point out that the first author Milena Schmidt is a doctoral candidate at the clinic for small animal medicine of the Ludwig Maximilian University in Munich. Furthermore, she is employed part-time at Napfcheck, a private small animal nutrition consultation practice in Munich with Julia Fritz as owner of the company (diplomate ECVCN). Besides nutrition counselling for dog and cat owners, the practice is also advising companies and giving lectures (f.e. for veterinarians, companies and pet owners). In the year 2015, J. Fritz published a book about the BARF-diet (“Hunde barfen: Alles über Rohfütterung”). Funding for analysis was provided from internal funds of the Gastrointestinal Laboratory at Texas A&M University. There are no other patents, products in development or marketed products to declare. This does not alter the objective and neutral perspective of the authors. This does not alter our adherence to all the PLOS ONE policies on sharing data and materials.


Fig 1
Fig 1. Intake of crude protein, fat, fiber, and NFE (Nitrogen-Free Extract) between both groups.
The BARF dogs were fed a significantly higher amount of protein and fat and lower amount of NFE and fiber.
Fig 2
Fig 2. Bacterial diversity measures.
(a) Alpha diversity: Rarefaction analysis of 16S rRNA gene sequences in raw and commercially fed dogs. Lines represent the mean of each group (red = BARF, blue = commercial), the standard deviation is shown by error bars. No significant difference in observed species was seen between BARF and commercially fed dogs. (b) Beta diversity: Principal coordinates analysis (PCoA) plot showing clustering of microbial communities from feces of raw and commercially fed dogs (red = BARF, blue = commercial). The closer the items, the more similar is the microbial community of the samples. Consequently, the microbiome of BARF dogs differs from the microbiome of commercially fed dogs (ANOSIM; p < 0.01).
Fig 3
Fig 3. Fecal abundance of selected bacterial taxa between BARF and commercially fed dogs based on qPCR.
The BARF dogs showed a higher abundance of Clostridium perfringens, E. coli and Streptococcus, while commercially fed dogs showed a higher abundance of Faecalibacterium. The Dysbiosis Index was significantly higher in the BARF group.
Fig 4
Fig 4. PCA plot of the fecal metabolome showing clustering of samples based on BARF versus commercially fed diets.
The BARF (red) and commercially (green) fecal samples were used to detect differences between the two diets. The PCA plot showed a clustering between both groups.
Fig 5
Fig 5. Heatmap of the most abundant metabolites of both groups, as identified by VIP scores in PLS-DA.
Every sample (10 BARF samples = red, 9 commercial samples = green) is represented by an own column, the more red colored the higher the abundance of these metabolites. The BARF dogs showed a higher abundance of e.g. myo-inositol, gluconic acid or isomaltose, the commercially fed dogs a higher abundance e.g. of catechol or phosphoethanolamine.
Fig 6
Fig 6. Random forest analysis.
Top 15 metabolites with the highest discriminatory power between both diet groups are listed. Red fields show a high abundance, green fields a low abundance of the particular metabolite based on diet.
Fig 7
Fig 7. Total, primary, secondary fecal bile acids and cholesterol of BARF and commercially fed dogs.
There was no significant difference in total, primary, and secondary bile acids between fecal samples of both groups, but BARF dogs had a higher abundance of cholesterol in their feces.

Similar articles

See all similar articles

Cited by 6 PubMed Central articles

See all "Cited by" articles


    1. Simon S. BARF: Biologisch Artgerechtes Rohes Futter. 5st ed. Wadern: Drei Hunde Nacht; 2012.
    1. Reinerth S. Natural Dog Food: Rohfütterung für Hunde-Ein praktischer Leitfaden. 1st ed. Norderstedt: Books on Demand; 2005.
    1. Dillitzer N, Becker N, Kienzle E. Intake of minerals, trace elements and vitamins in bone and raw food rations in adult dogs. Br J Nutr. 2011;106:53–56. - PubMed
    1. Isaiah A, Parambeth JC, Steiner JM, Lidbury JA, Suchodolski JS. The fecal microbiome of dogs with exocrine pancreatic insufficiency. Anaerobe. 2017;45:50–58. 10.1016/j.anaerobe.2017.02.010 - DOI - PubMed
    1. Vázquez-Baeza Y, Hyde ER, Suchodolski JS, Knight R. Dog and human inflammatory bowel disease rely on overlapping yet distinct dysbiosis networks. Nat Microbiol. 2016;1:16177 10.1038/nmicrobiol.2016.177 - DOI - PubMed

Publication types

Grant support

The authors received no specific funding for this work.