Raw feeding and the canine gut microbiome

What an eight-dog feeding trial found about raw beef, gut bacteria and stool quality, with a closer look at the methods, results and limits of the evidence.

Study summary · Dogs · Raw feeding · Sandri et al., 2017

Diet changed gut bacteria and stool quality.

In eight healthy adult Boxers, switching to a raw-beef mixed diet changed faecal bacteria and produced firmer stools. Bacterial diversity was higher than at the start of the study. These are short-term findings about diet and the gut microbiome; the trial did not establish that raw feeding produces better long-term health.

Participants8 adult Boxers
DesignRandomised crossover
Duration28 days
PublicationPeer reviewed · 2017

Why this study matters

Microbiome results are often used to support claims about pet food. This trial helps answer a narrower question: how much can faecal bacteria and fermentation products change when healthy dogs switch diets? It also shows why a measurable change needs careful interpretation before it becomes a claim about a healthier dog.

How much confidence should you place in this study?

The crossover design helps account for differences between individual dogs, but the lack of a washout period leaves possible carryover. Random allocation is reported; assessor blinding is not established. The model accounts for repeated measurements and reports some multiple-testing correction. Those strengths cannot overcome the very small, single-kennel sample or separate processing from recipe composition. Confidence is greater in a short-term microbiome change than in any claim of better health. Restricted underlying data access also limits independent checking.

Full-text editorial appraisal completed 23 September 2026. This assessment considers design, measurement, missing data, applicability and funding. It is not a formal certainty grade or a systematic review.

How the study worked

The dogs

Eight healthy adult Boxer dogs from a single kennel in north-east Italy: five females and three males, with a mean age of 4.2 ± 2.8 years. Starting mean weight was 30.3 ± 3.0 kg and body condition score was 4/9. Clinical examination and blood tests supported their healthy status.

The feeding trial

Dogs were randomly allocated to two groups of four. One group received the mixed diet for 14 days while the other continued with dry extruded food. The groups then exchanged diets for another 14 days, giving each dog both diets over 28 days. There was no gradual transition and no separate washout interval between the feeding periods.

The mixed diet contained approximately 70% raw beef by weight plus a purpose-made complement containing rice flour, chickpea flour, oat flakes, dried carrots, algal omega-3 fatty acids, vitamins and minerals. Water was added before feeding. The authors formulated it to NRC nutrient recommendations. This was a specific supplemented diet, not simply a bowl of raw meat.

What was used for comparison?

A commercial complete dry extruded diet. This was the food all dogs ate before the trial and the control food during it. The paper calls the starting observations the reference diet (RD), the same food during the trial the control diet (CD), and the experimental food the mixed diet (MD). These labels describe two foods assessed at different stages, not three different foods.

What did the researchers measure?

Faecal and blood samples were collected at baseline, day 14 and day 28. Researchers used 16S rRNA sequencing to describe bacterial groups, calculated diversity and evenness, measured faecal short-chain fatty acids and lactate, and assessed stool quality on a five-point scale. Weight, body condition and blood chemistry were also assessed. The authors defined stool scores of 2 to 3 as optimal, so a lower score is not automatically better across the whole scale.

What the researchers found

The table below separates the measured findings from what they can tell us about a dog’s health. Reported values with ± are means and standard deviations, showing variation within the measurements.

OutcomeReported resultHow to read it
Bacterial diversityShannon index: 3.31 ± 0.15 on the mixed diet, versus 2.92 ± 0.31 at baseline (P < 0.05). The control-diet value during the trial was 3.15 ± 0.09.The reported significant comparison was with the starting reference measurements. This is a diversity index, not a health score.
Stool qualityFaecal scores were lower on the mixed diet than at baseline (P < 0.05).Stools became firmer. This does not establish prevention or treatment of digestive disease.
Lactate and total short-chain fatty acidsLactate proportion increased, while total faecal short-chain fatty acids did not significantly change.The discussion distinguishes lactate proportion from absolute concentration. Faecal measurements are not a direct measure of production or absorption throughout the gut.
ButyrateThe proportion was numerically higher on the mixed diet, but the difference was not statistically significant (P = 0.081).This trial did not confirm an increase in butyrate. It should not be presented as a demonstrated benefit.
Weight and body conditionNo significant dietary effect. Mean weight was 30.1 ± 2.7 kg on the control diet and 29.9 ± 2.8 kg on the mixed diet.A short trial in healthy dogs cannot establish long-term effects on body weight or condition.

A P value below 0.05 was treated as statistically significant in the reported comparisons. Statistical significance does not measure the size of a health benefit. Equally, a non-significant result does not prove that two diets are equivalent.

A closer look at the microbiome results

The mixed diet changed the relative abundance of several bacterial groups. Lactobacillus and Paralactobacillus decreased (P < 0.01), and Prevotella was also lower (P < 0.05). Other groups, including Fusobacterium and Escherichia/Shigella, increased relative to baseline.

These results describe the proportions of bacterial DNA sequences found in stool. They do not establish that each increase was beneficial, that each decrease was harmful, or that an organism was causing disease. The study also found correlations between bacterial groups and fermentation products, but those associations do not prove a biochemical mechanism.

What strengthens the evidence?

The dogs were randomly assigned to the order of the diets, and each dog received both diets. This helps reduce the influence of differences between individual dogs. Housing and food preparation were controlled, and the raw meat came from one batch. Researchers combined bacterial sequencing with stool scores, fermentation measurements, body condition and blood tests. Their statistical model accounted for repeated measurements in the same dogs.

What limits the conclusions?

Only eight dogs were studied, all Boxers from one kennel. Each diet was fed for 14 days, which limits conclusions about other breeds, life stages, cats and long-term outcomes.

The diets differed in composition and physical form as well as processing. This was not a matched recipe tested raw versus cooked. There was no separate washout period, so carryover from the first diet is a concern in interpreting the crossover.

Bacterial diversity, stool scores and faecal fermentation products are intermediate outcomes. They cannot on their own demonstrate disease prevention, greater longevity or long-term feeding safety. Correlations between bacterial groups and fermentation products do not establish which organisms caused the changes.

The paper reports multiple microbial comparisons and some correction for multiple testing. The small sample still warrants caution. Underlying data were not openly available and required permission from Nutrigene.

What the researchers concluded

The authors concluded that diet composition changed faecal microbial communities and fermentation products. They interpreted the greater bacterial diversity and improved stool scores with the mixed diet as signs consistent with better gut health. Their discussion also acknowledged that establishing whether the microbiome changes represented better health was difficult.

Evidence.pet interpretation

This study supports the conclusion that a substantial diet change can alter the canine faecal microbiome over a short period. Firmer stools were a directly observed outcome, while lasting health benefits were not established. The comparison changed ingredients, nutrient composition, moisture and processing together, so it cannot isolate the effect of feeding meat raw. The findings do not justify a blanket recommendation for or against raw feeding, do not establish the safety of raw food handling, and cannot be transferred directly to cats, unbalanced homemade diets or a named commercial product.

Funding, author interests and data access

Nutrigene supplied funding and study materials. The work was supported within the regional Phytopet grant in partnership with the University of Udine. Bruno Stefanon was Nutrigene’s CEO, and Misa Sandri was a senior R&D scientist and Technical Manager. The company also controlled permission to access the underlying data. These interests are relevant when assessing the paper, but do not automatically invalidate its findings.

Read the original research

Sandri M, Dal Monego S, Conte G, Sgorlon S, Stefanon B. Raw meat based diet influences faecal microbiome and end products of fermentation in healthy dogs. BMC Veterinary Research. 2017;13:65. doi:10.1186/s12917-017-0981-z.

Evidence.pet editorial summary · Study ID STU-001 · Source checked 23 September 2026. Methods and results are summarised from the original paper. Strengths, limitations and practical interpretation are our editorial assessment, not a formal evidence grade. This summary does not assess any named retail product.

Adapted from Sandri and colleagues under CC BY 4.0. Wording has been condensed, reorganised and accompanied by Evidence.pet commentary.