A veterinary nutritionist checks a cat's diet against a short, fixed list of nutrients a cat's body cannot make enough of on its own — taurine, arginine, arachidonic acid, and preformed vitamin A — and then does the arithmetic that connects a package label to that list. Protein percentage never comes first, and on most bags it cannot answer the question at all. Those four nutrients are why a cat is an obligate carnivore. The number on the front of the bag is a rumor about them, not a measurement of them.
I learned that at a folding table in the back room of the rescue, surrounded by whatever food people had donated that month. No two bags alike. A kitchen scale I use for portioning, because a measuring cup is not a unit. My phone open to a calculator. And a board-certified veterinary nutritionist who had driven over on a Saturday, picking up the first bag and turning it face-down without reading the front of it.
"What's the moisture?" she asked.
Everyone in that room had come ready to argue about protein. That's what I did before that afternoon — on the internet, with strangers, over screenshots of ingredient panels. She asked about moisture because every number on the guaranteed analysis is given by weight, and weight is the one thing you cannot compare straight across a can and a bag.
The list she wrote on the back of the form
She turned the diet history form over and wrote four words. No formula, no brand. Then she drew a circle with an arrow leaving it, which turned out to be the urea cycle, and said that if we remembered one thing about feeding cats it should be that they're missing enzymes, not willpower.
Taurine comes first because cats lose it constantly. They conjugate most of their bile acids with taurine rather than glycine, so every pass through the gut drains a little more, and their own synthesis from cysteine and methionine is too slow to keep up. Heat processing takes another bite out of whatever was in the ingredients. So the diet has to deliver taurine preformed, every day, which is why it appears on labels at all. Its absence is tied to a specific form of heart disease — dilated cardiomyopathy — and to damage in the retina. This is the one item on the list most owners have heard of, and usually the only one.
Arginine is next, and it's the one that changes your sense of timescale. Cats run a urea cycle that depends on dietary arginine, and they don't have much ability to make it themselves. Ammonia is produced whenever protein is broken down, and the urea cycle is what clears it. What the nutritionist said twice, because she wanted it to land: this isn't a slow deficiency. A single meal that's short on arginine can push blood ammonia up within hours. That's why her checklist is a checklist and not a monthly average.
Arachidonic acid isn't protein at all, which is exactly why the protein argument misses it. Cats can't convert enough linoleic acid into arachidonic acid to meet demand, so they need it arriving preformed, from animal tissue. A food can post an impressive protein number and still be the wrong shape for a cat's fat metabolism. These are separate columns.
Preformed vitamin A is the last one, and the one where the direction of the check flips. Cats convert beta-carotene into retinol poorly, so the plant precursors that count for people and for dogs largely don't count here. Retinol from animal tissue does. And unlike most of the list, where the only question is whether there's enough, vitamin A has a ceiling as well as a floor. That asymmetry is a big part of why the list is diagnostic and not a recipe.
Then there's protein itself. Cats have a higher protein requirement than dogs, and they keep breaking amino acids down for energy even when fat and carbohydrate are supplying plenty of calories — the enzymes that do that don't shut off the way a dog's would. Protein is fuel for a cat, not just building material, which means the requirement is real and constant. It also means "how much protein" is a question about quantity, while the four items above are questions about whether the protein is the right kind.
The arithmetic that has to happen before you compare anything
Here's the part that embarrassed me, because it's just division.
The guaranteed analysis gives crude protein, crude fat, crude fiber, and moisture as percentages by weight. Take a canned food that's 10% protein and 78% moisture. The food part of that can is only 22% of it, so 10 divided by 22 gives you roughly 45% protein on a dry matter basis. Now take a dry food at 32% protein and 10% moisture: 32 divided by 90 is about 36%. The can looks like the higher-protein food.
Then correct for calories, because your cat eats calories, not grams. Say the can carries 1,050 kcal per kilogram — that works out to about 9.5 grams of protein per 100 kcal. Say the dry food carries 3,900 kcal per kilogram — about 8.2 grams per 100 kcal. The gap narrows considerably. Whether it flips depends on the actual foods.
I'm not telling you which one wins, because there isn't a universal winner. I'm telling you that if you compared those two foods by the number printed on the front, you compared nothing. Different moisture, different calorie density, different serving sizes. Two corrections, minimum, before the numbers mean anything.
What "complete and balanced" is actually certifying
The other thing I got wrong for years: I thought the label told me what was in the food. It doesn't. The guaranteed analysis lists minimums for protein and fat, maximums for fiber and moisture. Taurine shows up only if the maker chooses to list it. Arginine, arachidonic acid, and vitamin A generally don't appear at all.
So the label is the wrong document. The right one is the AAFCO nutritional adequacy statement, usually near the bottom in small type. It says one of two things. Either the food is formulated to meet the AAFCO Cat Food Nutrient Profiles for a named life stage, or a feeding trial substantiates that it provides complete and balanced nutrition for that life stage. Both are legitimate routes. They are not the same route — one is a calculation against a nutrient database, the other is months of feeding it to cats and measuring what happens — and it's reasonable to want to know which one you're buying.
Read the life stage carefully. Growth and reproduction is the most demanding profile. Adult maintenance is a lower bar. All life stages has to clear the growth bar. A kitten eating an adult maintenance food is eating something that was never designed for the cat in front of it.
And watch for the phrase intermittent or supplemental feeding. That is the label telling you, in the only place it's allowed to say it, that the food is not complete. It's a topper. It's fine as a topper. It's not a diet.
"High protein" means nothing, legally. There's no regulated definition, so every bag can claim it and none of them have to prove it. The measurable benchmark is the profile, for the life stage.
What she looked at that wasn't on the table
The food was maybe a third of the visit.
She put hands on the cat — ribs, spine, the muscle along the shoulder blades — and asked what he actually ate. Not what we put in his bowl. What he ate. The dog's bowl he walked past, the toppers, the treats, the cereal milk he'd get at if a glass was left down. She asked about the litter box and the water bowl. She asked whether the food changed when the foster house changed, which at a rescue it does, constantly.
That's the half of the check no label can do for you. The diet is what the cat ingests, not what you served. At a foster house with four cats on four different foods, that distinction stops being philosophical fast.
Where I still get it wrong, on purpose
I can read a label now. I can run the dry matter arithmetic in my head. I know the four words and why each one is on the list. I still don't formulate.
That's not humility, it's the actual conclusion the list leads to. The four nutrients are a set of reasons, and reasons are not a recipe. Getting taurine right in a food you cook at home means knowing how much the cooking destroyed and what the rest of the formula is doing around it. Vitamin A has a ceiling. Arginine has no room for error. This is a manufacturing problem before it's a kitchen problem, and the whole point of the list is that the margin is thin.
One more caveat, because people ask. Meeting the profile is a floor, not a ceiling. The profiles exist to prevent deficiency, and clearing them means the food is unlikely to hurt a healthy cat of the right life stage. It says nothing about whether it's the best choice for a cat with kidney disease, or diabetes, or a weight problem, or a cat who's pregnant. Those are different conversations with different answers, and some of them end with a diet that's deliberately restricted in one direction and adjusted in another.
So does a nutritionist check differently? Yes, and the difference isn't expertise in the abstract. It's that she started with four nutrients the cat can't build and worked outward to the label, while I'd spent years starting at the label and working outward to an argument. Same food. Same cat. Completely different order of operations.
The four words are in a note on my phone now. I don't use it to grade food. I use it to remember what I'm actually asking about when I ask whether a diet is complete.
