A Food Elimination Trial for a Cat: How to Run One That Gives an Answer

Quick answer

An elimination trial is a diagnostic test that happens to look like feeding. If it is run loosely, it produces no answer at all.

A food elimination trial is the only reliable way to diagnose food allergy in a cat. There is no blood test, saliva test or hair test that can do it, and the ones sold for this purpose do not perform well enough to act on.

The trial is also the step most often run badly. A single treat, a flavoured worming tablet, or one lick of another cat's bowl restarts the clock. Most trials that "did not work" were never actually run.

• Food allergy is real in cats, but it is not the most common cause of itch. Flea allergy is, and it has to be excluded first.
• The diet must be a strict novel protein the cat has genuinely never eaten, or a hydrolysed veterinary diet. Fish and chicken are almost never novel for a cat.
• Nothing else passes the cat's lips for the whole trial. No treats, no dairy, no flavoured medication, no dental chews, no scraps.
• Skin signs usually need eight weeks or more before you can judge. Gut signs often improve faster.
• The trial is only a diagnosis if you rechallenge with the old food at the end and the signs come back.

Decide in 60 seconds

Why a trial has to be this strict

Food allergy is an immune response to a protein the cat has been eating, often for months or years. Reaction size does not depend on portion size in any useful way.

That is the whole reason for the strictness. A trace amount of the offending protein can be enough to keep the skin inflamed, and if the skin never fully settles you cannot tell whether the diet is working.

Inflammation in skin is also slow to resolve. Even after the trigger is completely removed, the itch, the scabs and the regrowing hair take weeks to catch up. A trial stopped at three weeks because "nothing has changed" was stopped before the test had finished running.

Gut signs behave differently. Vomiting and stool consistency often improve within two to three weeks, which is why a cat with gut signs only can sometimes be assessed sooner.

Rule out the more common causes first

Fleas.

Flea allergy dermatitis is the single most common cause of itch in cats, and a cat that grooms compulsively removes the evidence. Not finding fleas on a cat is close to meaningless.

Before you spend two months on a diet, put every cat and dog in the household on a veterinary flea product, used at the correct interval, without a gap, and treat the environment. If the itch resolves, you have your answer and you have saved yourself the trial.

Mites and other parasites.

Ear mites, Cheyletiella, Notoedres and Demodex all cause itch that mimics allergy. This is what the skin scrapes and ear swabs at the first appointment are for.

Pain and stress.

Cats overgroom over a painful area, and they also overgroom as a displacement behaviour. A cat with a bald belly may have bladder pain or arthritis rather than any allergy at all. This is a genuine and frequently missed alternative, and a diet trial will not touch it.

Environmental allergy.

Cats react to pollens, house dust mites and moulds much as dogs and people do. It is seasonal in some cats and year-round in others, and a food trial is partly how it gets identified: if strict feeding changes nothing, the environment moves up the list.

Choosing the trial diet

Weigh the daily ration out at the start of the day. It is the only way to know at the end of the day whether the cat actually ate.

There are two workable options, and your vet will help you pick between them.

A strict novel protein.

A protein this individual cat has genuinely never eaten. That means going through every food, treat and table scrap the cat has ever had, including what was fed at the shelter or by the previous owner.

Chicken, fish, beef and dairy are the proteins most often implicated in cats, and they are also in almost every commercial cat food, which usually rules them out as novel. Fish in particular is not novel for most cats even if it has never been the headline ingredient.

A hydrolysed diet.

The protein is broken into fragments small enough that the immune system does not recognise them. This is the more reliable route when the cat's dietary history is unknown, which is common in rescue cats.

Whichever route you take, use a veterinary formulated product. Supermarket "limited ingredient" foods have been found repeatedly to contain undeclared proteins, and a contaminated trial diet gives a false negative that sends everyone down the wrong path.

Do not home-cook the trial diet unless a veterinary nutritionist has formulated it. A meat-only home diet run for eight weeks creates a serious calcium and mineral imbalance of its own, and in a young cat that is a bone problem rather than a skin one.

Running the trial without contaminating it

In a multi-cat household the trial fails at the second bowl. Feed separately, in separate rooms, and pick the bowls up afterwards.

Transition gradually.

Cats form strong food preferences and a sudden switch often means a cat that will not eat. Mix a small proportion of the new diet into the old over seven to ten days, increasing slowly. Warming wet food slightly increases its smell and helps.

Weigh out the day's ration.

Put the whole daily amount in a container in the morning and feed from it. At the end of the day you know exactly what went in, which matters both for the trial and for spotting appetite loss early.

Remove every other food source.

No treats, no dental chews, no cat milk, no dairy, no scraps, no licking plates, no flavoured toothpaste, no flavoured supplements or joint powders.

Check the medication.

This is where most trials quietly fail. Flavoured worming tablets, flavoured parasite chews, palatable pain medication and flavoured probiotic sachets often contain animal protein. Ask your vet to switch to unflavoured equivalents for the duration.

Separate the other pets.

Feed the trial cat alone, in another room, and pick the bowl up. Move the dog's food and the other cats' food out of reach, and remember that dog food left down all day is a trial killer.

Restrict outdoor hunting and scavenging if you can.

A cat that hunts or visits a neighbour who feeds it cannot be trialled reliably. If neither can be prevented, say so at the start rather than discovering it at the end.

Keep flea prevention going throughout.

Stopping flea control during a food trial guarantees an uninterpretable result.

Score the itch weekly.

Give it a number out of ten each week, in the same conditions, and photograph the worst affected area. Slow improvement is invisible from memory and obvious from photographs.

Reading the result, and the step people skip

At the end of the trial period, compare your weekly scores and photographs, not your impression.

If nothing has changed at all, and you are confident the trial was clean, food allergy is unlikely and the investigation moves toward environmental allergy or another cause.

If the signs have clearly improved, you are not finished. You have a cat that improved while several things changed at once, including the season, the flea treatment and the amount of attention it was getting.

The confirmation step is the rechallenge. Go back to the original food and watch. If the itch or the gut signs return, usually within days to a couple of weeks, the diagnosis is made.

That step feels cruel and owners often refuse it. It is worth doing, because it is the difference between knowing your cat has a food allergy and committing to an expensive prescription diet for a decade on a hunch.

Once the reaction is confirmed, your vet may reintroduce single proteins one at a time, several weeks apart, to find which one the cat reacts to and widen the choice of long-term foods.

What never to do

Do not use a blood, saliva or hair test to diagnose food allergy and skip the trial. These tests do not reliably distinguish allergic from healthy cats, and acting on one means a diet change based on nothing.

Do not switch to a grain-free food and call it an elimination trial. Grain allergy in cats is uncommon, and grain-free foods are usually full of the same meat proteins the cat was already eating.

Do not run the trial and the flea treatment change and a new medication all in the same week. If several things change together you cannot attribute the improvement to any of them.

Do not give steroids during the trial without discussing it first. They suppress the itch regardless of cause and make the result impossible to read, though there are cases where a vet will deliberately use them early for welfare reasons and time the trial around it.

Do not give a cat any human antihistamine or anti-itch product from a cupboard. Dosing is species specific and several human skin and pain products are toxic to cats.

Do not treat a cough or wheeze as an allergy problem to be solved by diet. Coughing cats need their chest assessed, and feline asthma is treated in a completely different way.

When to get help

A trial that is run cleanly ends with either a clear answer or a clear reason to look elsewhere. A trial run loosely ends with neither.

See a vet the same day if the cat stops eating, becomes lethargic, develops sudden facial swelling, or starts breathing with effort or with an open mouth.

Book an appointment before starting a trial rather than after. The skin scrapes, ear swabs and parasite treatment that come first are what stop you spending two months testing the wrong hypothesis.

Book sooner than planned if the skin becomes broken, weeping, crusted or smells, because self-inflicted wounds become infected and that needs treating on its own.

Bring the food history, the weekly itch scores, the photographs, the exact product names of everything the cat is given including medication, and the start date of the trial. A dermatology referral is worth asking about if a clean trial and good flea control have both failed to explain the itch.