Fifty-one raw-fed dogs walked into a vet clinic and came out with better waistlines than the kibble group.

A study published in BMC Veterinary Research on February 6, 2026 compared 104 healthy dogs: 51 eating a raw meat-based diet and 53 eating a commercial complete diet. Each dog got two clinical evaluations three to five months apart. The raw group's body condition scores landed exactly where veterinarians want them to land.

That half of the paper is a clean win.

The same workup told a second story.

Four minerals came back low in the raw-fed dogs. The calcium-to-phosphorus ratio in their meals was off target. Both findings deserve to be said plainly, so here they are.

Neither one is a reason to quit raw feeding, and both are worth fixing this week.

Body condition score runs on a 1 to 9 scale, where 4 and 5 are ideal.

The raw-fed dogs had a median score of 5, with most of them falling between 4 and 5. The commercial-fed dogs had a median of 6, a full point above ideal, with the upper range reaching 7. The gap carried a p value below 0.001, which means it is very unlikely to be noise.

One group sat in the textbook range and the other sat above it.

A single point on that scale is not cosmetic.

Scores above 5 are how a veterinarian flags a dog drifting toward overweight. The commercial group's median put them in that zone at a routine checkup. The raw group's median did not.

That is the kind of difference an owner can see from across the room.

The energy numbers make the body condition result stranger.

The raw rations supplied 89 percent of each dog's daily recommended metabolizable energy. The commercial diets supplied 102 percent. The difference held up statistically at p = 0.015.

The raw-fed dogs were eating under the recommendation and holding ideal condition anyway.

Feeding math on a fresh diet rarely matches the number printed on a bag.

Owners weigh, portion, and adjust by eye when they build meals themselves. Those small corrections stack up over months. The data caught a group that had quietly settled below the official target and looked better for it.

Portion control gets easier when the food is not engineered to be overeaten.

Serum iodine was significantly lower in the raw-fed dogs at p = 0.001.

Copper, zinc, and manganese were also reduced in that group, each at p < 0.05. Those four minerals handle unglamorous jobs: thyroid signaling, skin and coat, immune function, and enzyme activity. None of them are optional over a long life.

Low is not the same as deficient, and the distinction carries real weight here.

Iodine is the one to watch closest.

It produced the strongest signal of the four minerals at p = 0.001. Thyroid function depends on a steady supply of it. Meat, organ, and bone carry very little iodine on their own.

A raw bowl can miss iodine almost entirely without anyone noticing for months.

No raw-fed dog in the study showed clinical signs of nutrient deficiency.

Coats were fine. Physical exams were fine. The shortfall showed up in bloodwork and in the ration math, not in the dogs themselves.

This is a margin problem caught early, not an emergency.

Calcium and phosphorus work as a pair, and the ratio between them matters more than either number alone.

The raw rations in this group ran a calcium-to-phosphorus ratio of 1.0 against a recommended 1.4. Phosphorus was keeping pace with calcium at a point where calcium should be comfortably ahead. The p value came in below 0.001.

Muscle meat is loaded with phosphorus and nearly empty of calcium, which is exactly how a ratio ends up flat.

A bowl of ground beef and chicken breast is not a raw diet.

It is a protein source missing the parts that balance it. Bone brings the calcium. Organs bring the trace minerals.

Leave those out and the ratio tips, every time.

The honest read on this study is narrow and specific.

Home-built raw rations in this cohort had gaps. A raw diet assembled well does not have those gaps. The variable sitting in the middle of this result is the recipe, not the category.

Nobody looks at a badly built salad and decides vegetables are the problem.

Kibble closes these same four gaps a different way.

Extrusion is hard on nutrients, so manufacturers add synthetic nutrients back in as a premix after the heat has done its work. The bag then meets the nutrient profile on paper. The base underneath that premix is still an ultra-processed product.

Fortification papering over an ultra-processed base is a trade-off, not a fix.

Every shortfall the paper found has an answer that comes from food.

Liver and other organ meats carry copper and zinc in forms dogs absorb well. Raw meaty bones or a ground bone source bring the calcium that pulls the ratio back up toward 1.4. Kelp is a genuine iodine source, and a small measured amount goes a long way.

Manganese shows up in organ meat as well, so one fix quietly covers two of the four.

The fourth fix is arithmetic.

Run the recipe past someone who can check it against a published nutrient profile. A veterinary nutritionist or a well-built formulation tool will catch an inverted calcium-to-phosphorus ratio in minutes. The authors of the study recommend veterinary nutritional assessment for dogs on raw diets.

Agreeing with that recommendation costs raw feeders nothing.

Raw feeders tend to resist the idea that their food needs checking.

The resistance is understandable after years of being told fresh food is reckless. It also leads straight to the gaps this paper measured. A recipe that gets audited once a year is the version of raw feeding that holds up under a blood panel.

Two minutes with a spreadsheet protects the whole argument.

Put both halves of this result side by side and the pattern is hard to miss.

Raw feeding got the hard part right without appearing to try, the part the pet food industry has been failing at for decades. Dogs at ideal body condition, eating less food than the guidelines call for, with no clinical signs of deficiency anywhere in the group. The part raw feeding got wrong is a shopping list and some fifth-grade math.

One of those problems takes a weekend to fix, and the other one is baked into the bag.