The 8-week elimination diet trial: every detail that affects the result

The elimination diet trial is the gold standard for diagnosing cutaneous adverse food reactions in dogs and cats. It is also the most often mis-run test in veterinary dermatology. The number of trials I see that produced an "inconclusive" result because someone gave a treat, used a flavoured medication, or fed a different protein — and that could have been conclusive if those small things had been caught — is depressing.

This is the strict version that gives a clean answer. Most of the article is the protocol. The end has the most common ways the trial fails.

What an elimination diet trial actually does

The trial has two phases:

Phase 1 (weeks 1–8): strict elimination. The dog or cat eats only one novel protein + one novel carbohydrate source for 8 weeks. No treats, no table food, no flavoured medications, no dental chews, no supplements that contain protein. The goal is to remove every possible food allergen from the diet.

Phase 2 (weeks 9–12): provocation. Ingredients are reintroduced one at a time, one per week, to identify which ingredients trigger a reaction. A reaction is typically visible within 3–7 days of reintroduction.

The trial takes 12 weeks total minimum. The 8-week strict phase is the part that is most often compromised.

The protein selection

There are two schools of thought, and the choice matters:

Novel protein diet

A protein the animal has never eaten before. Common choices for dogs: venison, rabbit, duck, kangaroo, alligator. For cats: rabbit, venison, duck, novel fish (sardine, mackerel).

The challenge: many "novel" proteins have been fed to pets in commercial diets, treats, or table scraps. A dog that has eaten chicken-based kibble for years and is "allergic to chicken" can be tested with rabbit — but if the dog has eaten rabbit treats in the past, rabbit is no longer novel.

Solution: review the pet's full dietary history with the owner before choosing. Any protein mentioned in the history is not novel.

Hydrolyzed protein diet

The protein is broken down into fragments too small for the immune system to recognise. Common options: Hill's z/d, Royal Canin Hypoallergenic, Purina Pro Plan HA, Royal Canin Anallergenic.

The challenge: not all hydrolyzed diets are created equal. The molecular weight of the protein fragments must be under 10 kDa to reliably avoid immune recognition. Anallergenic uses feather protein hydrolysate at very low molecular weight; z/d and HA use chicken or soy. Some dogs with severe chicken allergy still react to poorly-hydrolyzed chicken diets. If a trial fails on one hydrolyzed diet, switching brands sometimes gives a clean answer.

For most dogs, I use a hydrolyzed diet unless the dietary history clearly suggests a novel protein that hasn't been fed.

The strict list of what is and isn't allowed

Allowed

• The selected protein source (fresh-cooked or commercial hydrolyzed)

• The selected carbohydrate source (sweet potato, potato, oats, rice, or the carbohydrate that comes with the hydrolyzed diet)

• Water

• The elimination diet itself, in any form (kibble, canned, freeze-dried, rehydrated)

Not allowed

• Treats — none, not even single-ingredient treats. The protein in the treat may be the very allergen you are trying to identify.

• Table food — none. No small piece of cheese, no sample of what you are eating, no licking of plates.

• Flavoured medications — heartworm preventives, joint supplements, probiotic pastes, and many antibiotics come in flavoured formulations. Switch to unflavoured versions during the trial.

• Flavoured toothpaste (for pets that tolerate brushing) — use unflavoured or skip brushing during the trial.

• Dental chews — most contain hidden proteins.

• Toothpaste / oral gels with animal-derived flavouring.

• Nutraceuticals / supplements unless explicitly cleared with the vet.

• Topical products with animal-derived ingredients — some paw balms contain lanolin or beeswax.

• Access to other pets' food — keep cats separated from other-cat food; keep dogs out of the kitchen during human meal prep.

This list is long because the trial is strict. Every item on the "not allowed" list has caused an inconclusive trial at some point in the literature.

Week-by-week timeline

Week 1: Transition week

If switching from a previous diet, mix the elimination diet in increasing proportions over 7 days:

• Days 1–2: 25% new diet, 75% old

• Days 3–4: 50/50

• Days 5–6: 75% new, 25% old

• Day 7: 100% new diet

Some pets develop loose stool during transition. This usually resolves in 3–5 days. If it persists past day 7, contact the vet.

Weeks 2–4: Establishing baseline

The pet's clinical signs (itch, ear infections, GI upset) should start improving during this window if food is the trigger. Improvement is gradual — owners often don't notice a change until week 4–5.

What to monitor daily:

• Pruritus score (0–10 scale)

• Stool quality

• Ear inflammation

• Skin lesions

• Energy level

Weeks 5–8: The strict phase

If the pet has improved, continue the strict diet. If the pet has not improved at all by week 8, food is unlikely to be the cause and the trial can be discontinued (the pet can return to a previous diet).

Partial improvement (some signs better, others not) is common. Food allergy frequently coexists with environmental allergy (atopic dermatitis). In that case, the trial is still worth completing to identify the food component contribution.

Week 9: First provocation

Add a single protein back into the diet. Continue the elimination diet as the base, but add one novel protein (chicken is the most common provocation, since it's the most common allergen).

• Days 1–3: small amount of the test protein mixed into the elimination diet

• Days 4–7: increase the proportion

• If no reaction in 7 days, that protein is cleared

Weeks 10–12: Subsequent provocations

Continue adding proteins one at a time, one per week. The order is typically:

1. Chicken (most common allergen)

1. Beef

1. Dairy

1. Wheat

1. Egg

1. Soy

1. Lamb

Each protein: 7 days of provocation, then return to elimination diet for 3 days washout, then move to next. If a reaction occurs (return of itch, GI upset, ear inflammation), that protein is the allergen — avoid it permanently.

A dog that has no reaction to all 7 common allergens has either a non-food allergy (most likely atopic dermatitis) or a rarer allergen not in the test list. Further workup follows.

How to know the trial is working

Three signs during weeks 4–8 indicate the trial is producing useful data:

1. Pruritus drops by 50%+ — strong indication that food is a contributor.

1. Ear inflammation resolves — chronic ear infections that improve dramatically during the trial point strongly at food.

1. GI signs normalise — chronic soft stool, intermittent vomiting, or flatulence that resolves during the trial.

If none of these happen by week 8, food is unlikely to be a major contributor. Move on to other diagnostics.

Common ways the trial fails

A list of the failures I see in clinical practice:

1. The "small" treat

"I only gave him one tiny piece of cheese." Cheese is dairy. Dairy is one of the most common canine food allergens. One piece is enough to trigger a flare in a dairy-allergic dog and ruin the trial for that week.

2. The flavoured pill

The dog is on a flavoured heartworm preventive. The flavoring contains animal protein (often pork or beef). The trial is being sabotaged daily.

Fix: switch to unflavoured tablets or topical preventives for the duration of the trial.

3. The other-pet food

Two-cat household. One cat is on the elimination diet. The other cat is on regular food. The first cat eats the second cat's food when the owner is not looking.

Fix: separate feeding locations, separate feeding times, supervise all meals.

4. The hidden protein in supplements

The dog is on a glucosamine/chondroitin supplement. The chondroitin is often shark or bovine cartilage. It contains protein.

Fix: clear every supplement with the vet. Replace as needed.

5. The non-compliant family member

The spouse or roommate gives the dog treats "because he looks hungry." This is the single most common cause of failed trials.

Fix: a household meeting before starting the trial. Everyone agrees to the protocol. The vet can provide written instructions.

6. The "all natural" chew

"Bully stick is just meat. It must be fine." Bully sticks are beef. Beef is one of the common allergens. The trial is sabotaged.

Fix: no chews of any kind during the trial. If the dog needs to chew, use the elimination diet kibble as chew rewards.

7. The unannounced ingredient change

The owner buys the same brand of hydrolyzed diet as last time, but the manufacturer has changed the formula. The new formula has a different protein source.

Fix: check the ingredient list every time you open a new bag. Manufacturers do change formulations without notice.

The bottom line

The elimination diet trial is the gold standard for food allergy diagnosis, but only if it is run strictly. The 8-week strict phase requires removing every possible allergen from the diet, including treats, flavoured medications, dental products, and supplements. The provocation phase takes another 4 weeks to identify which specific ingredients are the triggers. Most trials that produce inconclusive results did so because of one of the seven common failures above. Strict execution is the difference between a clean answer and a wasted 12 weeks.

---

About the author: [Dr. Hannah Kim](../Hannah%20Kim.md), DVM, Cert. VD, is a Seoul National University-trained dermatologist with 11 years of practice, and currently the director of Pacific Pet Dermatology in Vancouver.