Dog Food Nutrition and Feeding Trials


I received the following message on WDJ’s Facebook page:

Hey WDJ! My wife and I love your website/magazine and constantly link it and recommend it. Keep up the great work!

I was hoping to have a discussion with you about this article:

How do you reconcile that feeding trials are the best way, imperfect as they are, to determine if a food is nutritionally sound, with the fact that hardly anyone (or their foods) does AAFCO feeding trials? My favorite dog food companies (aka the ones I learned about through your yearly reviews, like Merrick) don’t seem to have any AAFCO feeding trials under their belt.

I’m not trying to question your knowledge or recommendations (because I love both), more just hoping I’m not the only one out there thinking about this stuff so much. 

Thanks for your question! I’ve discussed feeding trials in a lot of different articles over the years, and I’m happy there are others “thinking about this stuff” as avidly as I am.

It’s true that in the article you mentioned in your note, I said that the “nutrient levels” method for establishing the nutritional adequacy of a food was flawed:

“Feeding trials are considered by most veterinary nutrition experts [emphasis added] to be the ‘gold standard’ for proving nutritional adequacy claims – superior to the ‘nutrient levels’ method of proof. That’s because it’s quite possible for a laboratory analysis to confirm that a food contains the amounts of various nutrients judged to be necessary for maintaining a dog, but for the product, in practice, to fail at that very job.

This is possible because not all nutrients may be in a digestible (‘bioavailable’) form. Most nutritionists agree that feeding trials offer the most reliable confirmation of a food’s ability to deliver nutrients in a form that will benefit the target species.”

However, as I’ve mentioned in many of my other articles on commercial diets, the feeding trial method of establishing nutritional adequacy is flawed, too.

The above-referenced article didn’t assert that feeding trials are the best way to prove the nutritional adequacy of a food. Its purpose was to clarify what feeding trials are, and describe conditions for dogs used in the studies.

In that article, I did call feeding trials “important” – and I still think they are. It absolutely is important to know whether a food that may well be fed to a dog every day for years on end is, in actuality, capable of sustaining dogs over time, without causing gross deficits leading to illness, weight loss, or abnormal blood chemistry.

I’m not going to go so far as to regard them as requisite, however, because they aren’t perfect. For one thing, they really aren’t long enough. Just because a diet can sustain a dog in a laboratory environment for about six months without causing illness or abnormal blood values doesn’t mean it will perform the same way for dogs who may lead a much more active and stressful life, and for years on end.

Also, as I explained in my 2007 dry food review (
), foods that acquire the right to use a nutritional adequacy claim based on feeding trials need not be formulated to meet the other standard for nutritional adequacy: the “nutrient levels” criteria. Here is an excerpt from the 2007 article – but I’m going to boldface and correct a big mistake I made there:

“Foods that pass feeding trials are not required to contain minimum or maximum levels of any particular nutrients. Therefore, it’s possible for a food to sustain dogs long enough to ‘pass’ the trial, but fail to demonstrate an ability (in real-world, long-term use) to promote optimum health. As one example, mineral excesses may take a year or more to cause noticeable health problems, but a food that claims to provide complete and balanced nutrition for adult dogs (a ‘maintenance’ claim) may have passed only a 26-week test.

“There is also an AAFCO feeding trial (at least 13 weeks long) for products intended for dogs during gestation and lactation and another that tests puppy diets (10 weeks). To earn the right to claim nutritional adequacy for dogs of ‘all life stages,’ a food must undergo all three trials sequentially, for a total of 49 (or more) weeks. [Actually, to earn the ‘all life stages’ claim, the food must pass the ‘gestation and lactation’ and then the ‘growth’ (puppy) trials, sequentially, for a total of about 23 weeks. The ‘maintenance’ trial is not actually included.] If it passes, its label can state, ‘Animal feeding tests using AAFCO procedures substantiate that (name of product) provides complete and balanced nutrition for all life stages.’

“Many experts regard the ‘all life stages’ feeding trial as the best proof we have of a food’s ability to perform. But again: Even a year-long [nope, only six-month!] feeding trial may fail to reveal faults that can cause serious health problems if fed as a sole diet for a long period.”

However, as I said in the 2007 article, the “nutrient levels” claim is flawed, too:

“Foods that meet the ‘AAFCO nutritional profiles’ qualification can lack palatability and/or digestibility. If dogs don’t like the smell or taste of the food, they won’t eat enough of it to gain its nutritional benefits. Also, the nutrients contained in a product may not be present in a form that the dog can digest. The AAFCO nutrient profiles themselves contain a problem: Not many lay people are aware that the profiles allow for a wide range of values. Far from being some sort of industry ‘standard,’ or offering suggestions for optimum nutrition, they actually offer only broad guidance.”

The fact is, both methods that a company can use to “prove” the nutritional adequacy of a product are flawed. Here’s an excerpt from a 2012 article I wrote about the research conducted by pet food companies (

“Many people consider AAFCO feeding trials as the ‘gold standard’ for confirming the nutritional adequacy of a canine or feline diet. However, because the products that met the ‘feeding trial’ test of nutritional adequacy do not have to meet the ‘nutrient levels’ criteria, there is the possibility that these products may contain excessive, deficient, or unbalanced nutrient levels that may contribute to the development of health problems if fed as a sole diet for periods that are in excess of the testing period. An ideal test would encompass both a feeding trial and meeting the AAFCO nutrient profiles, but no such standard currently exists [emphasis added].”

Veterinarians are taught in vet school (with texts underwritten by pet food companies) that only foods qualified by feeding trials should be fed, ever. And the vast majority of veterinarians believe this. Maybe it’s because of all the blood tests a dog in a feeding trial is subjected to . . . But six months! It’s not enough to base a lifetime of feeding on, in my opinion.

That’s why I don’t say in our pet food reviews that one ought to use the type of nutritional adequacy claim as a selection criteria. I do think, however, that this information is worth knowing –that dog owners should always keep in mind which test was used to prove the adequacy of their dogs’ food as they monitor their dogs’ health and condition closely. If it’s a “feeding trial” product – ask the maker (or better yet, look for yourself) to find out if it DOES meet the “nutrient levels” standards, or do some nutrient values deviate from the AAFCO Canine Nutrient Profiles? If it’s a “nutrient levels” product, ask the maker what sort of informal feeding trials they use, how long the diet is fed to its test dogs, and what sort of tools are used to monitor or evaluate the dogs used in the trials. Some companies use their employees’ dogs or the dogs in a shelter close to the company headquarters as informal test dogs, but don’t follow up with any sort of health tests. These informal tests really only give the company information about the palatability and digestibility of the product; they don’t address long-term health consequences. But then, neither do the AAFCO-protocol feeding trials, unless you consider six months to be “long term.”

— Nancy Kerns, WDJ Editor