Credit: Kibble Facts
A team at the Universidad de Chile bought ten kibbles and six raw meat-based diets, ran them through a lab, and then sequenced the gut bacteria of 24 healthy dogs eating them. The study ran in Frontiers in Microbiology on September 25, 2026. Most of the attention went to the bacteria. The number that explains the bacteria is in the diet table. The kibble carried nearly five times the carbohydrate of the raw food, and the gut microbiome of the dogs eating it looked different in ways the researchers could measure.
What the researchers fed and measured
This was a cross-sectional look at clinically healthy dogs, not a feeding trial. The team did two things at once.
Proximate analysis of 16 commercial diets, 10 extruded and 6 raw meat-based, measuring protein, fat, fiber, and carbohydrate on a dry-matter basis.
16S rRNA sequencing of fecal samples from 24 dogs: 14 eating extruded kibble and 10 eating raw.
The dogs were household pets on whatever their owners already fed. Three were puppies, and the team re-ran the analysis without them to check the result held. It did. The sample is small, which the authors say plainly, and the design cannot prove the diet caused the bacterial differences. What it can do is show two groups of dogs eating two very different things and carrying two different communities in their gut.
The carbohydrate gap
Here is what the diets contained per 100 grams of dry matter.
Nutrient | Raw meat-based | Extruded kibble |
|---|---|---|
Crude protein | 51.2 g | 26.8 g |
Fat | 28.9 g | 12.7 g |
Non-fiber carbohydrate | 10.2 g | 48.9 g |
Crude fiber | 1.1 g | 3.4 g |
Nearly half of the kibble, by weight, was starch and sugar. The raw diets were about a tenth carbohydrate and roughly double the protein. That is not a tweak. It is a different food aimed at a different digestive system. A dog's digestive physiology is built around protein and fat, and the kibble recipe is built around what an extruder needs to form a pellet. Starch is the binder. The dog is the one who eats the binder.
What changed in the gut
The sequencing results line up with the diet table.
More kinds of bacteria in raw-fed dogs. Chao1 richness was higher in the raw group (p = 0.04). The Simpson index also differed (p = 0.02).
A different community structure. Weighted UniFrac, which compares the overall makeup of the bacterial population, separated the two groups (p = 0.003).
Different dominant groups. Fusobacteriota made up 39.8% of the raw-fed community. Bacillota, the phylum formerly called Firmicutes and a heavy fermenter of starch, made up 43.4% of the kibble-fed community.
Fusobacteria rise when dogs eat meat. Earlier work from an Italian team found the same shift in a controlled study where dogs were switched from kibble to raw and their fecal bacteria changed within weeks. The Chilean results are the household version of that finding. Feed a dog meat and you get a meat-eater's gut. Feed it starch and you get the bacteria that live on starch.
What the paper cannot tell you
A few honest limits.
Twenty-four dogs is a small group. Several measures of diversity did not reach significance, and the differences at the level of individual bacterial species did not survive the statistical correction for multiple testing.
Cross-sectional means a snapshot. The team did not switch any dog's diet and watch what happened. The Italian trial above did, and found the same direction.
Six raw diets is a thin sample of a big market. All six deviated from their label in at least one nutrient, against one of the ten kibbles. That is a reason to buy raw from a company that publishes its own lab results, not a reason to buy kibble.
None of those limits touch the diet table. The carbohydrate numbers are from the lab, not from the dogs. A kibble that is 48.9% starch and sugar by dry weight is 48.9% starch and sugar whether you sequence anything or not.
The pattern
Every time researchers put raw-fed and kibble-fed dogs side by side, the same two things show up. The raw diet carries a fraction of the carbohydrate, and the raw-fed gut carries a different, usually richer, set of bacteria. The 2026 Chilean paper adds a label analysis that most papers skip and finds the raw sector has homework to do on accuracy. The fix for that is choosing verified raw brands that test and publish. The fix for a bag that is half starch does not exist, because the starch is what holds the bag's contents together.
Sources
Solís D, Toro M, Navarrete P, Medina DA, Reyes-Jara A. "The fecal microbiota of healthy dogs fed extruded and raw meat-based diets." Frontiers in Microbiology, Vol. 17, September 25, 2026. https://www.frontiersin.org/journals/microbiology/articles/10.3389/fmicb.2026.1910251/abstract — Peer-reviewed. Observational, cross-sectional. Small sample (24 dogs). No conflicts identified.
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. https://pubmed.ncbi.nlm.nih.gov/28245817/ — Peer-reviewed controlled feeding trial. Small sample.
