Pet Food Company On Trial

The truth about pet food feeding trials.


[Updated March 4, 2016]

In 2003, People for the Ethical Treatment of Animals (PETA) launched a media campaign aimed at exposing what it characterized as the cruelty, pain, and suffering experienced by dogs and cats that are used in laboratory research studies conducted by pet food companies.

One of the more unfortunate results of this campaign, as far as we’re concerned, is that consumers may have become convinced by PETA’s campaign that two important lab tests – feeding trials and metabolic studies – are inherently cruel. In fact, there is nothing innately cruel about the tests themselves; neither imposes anything more invasive than the removal of a blood sample on the canine test subjects. Also, we believe the tests do benefit the canine population at large. In our opinion, it’s how (and how long) the dogs are kept by the labs that could stand major improvement.

Pet Food Company On Trial

There are all sorts of other laboratory research studies conducted by the larger pet food manufacturers that push the envelope of “humaneness.” Some companies perform (or order a “contract lab” to perform) studies that call for a disease or health problem to be induced in a population of test dogs, such as a damaged kidney or cardiac problem. Then various nutritional approaches to improving the condition are tried. Some of these tests also require euthanasia and postmortem studies of the dogs; feeding trials and metabolic studies do not.

Research conducted to formulate foods that improve health problems has the most potential for inflicting pain and suffering on the test dogs. But it shouldn’t mistaken for what happens in a feeding trial to confirm the nutritional adequacy of a food – which is what this article is about. We’ll save for another day a discussion about the ethics and practice of conducting research that requires the induction of disease and euthanasia of the test subjects.

Definition of a Feeding Trial

It should not come as a surprise that pet food makers use dogs and cats to develop and test their products, and not just in a “He likes it! Hey Mikey!” kind of way. The most common test ordered by dog food makers is the “feeding trial,” a formal test where dogs are fed a certain food and nothing but that food for about six months. Feeding trials are conducted to test and – it is hoped – establish the ability of a food to keep dogs alive and reasonably well. When a food “passes” a feeding trial, its maker earns the legal right to include a statement on the product label assuring consumers that a feeding trial has established the nutritional adequacy of the food.

The “metabolic study” is another increasingly common test used by pet food makers. It requires that a dog spend at least five days in a relatively uncomfortable cage with a slatted metal floor, which allows technicians to collect every bit of the subject’s urine and feces. We’ll describe it in more detail below, but suffice to say it’s a common test, not invasive, but not particularly fun for the dog, either.

Sound like a case for PETA? The high-profile animal rights organization opposes all laboratory testing on animals, so it was perhaps just a matter of time before it trained its sights on the laboratory testing conducted by pet food makers.

However, PETA’s metaphorical weapon isn’t sharpshooter-precise; it tends to be more like a shotgun splatter of allegations of animal cruelty. PETA has taken the pet food industry (with a special concentration on one unlucky company in particular) to task for the all the research it conducts to develop and establish foods. (For more on the PETA campaign, see “Truth Is the First Casualty of War” sidebar.) But again, we’re going to limit our discussion to just two pet food tests that we feel are useful.

“Complete and Balanced”

Go grab a can or bag of your dog’s food, and look for a little bit of fine print that mentions AAFCO – the Association of American Feed Control Officials. The statement tells you what sort of proof the manufacturer used to establish the nutritional adequacy of that product. Pet food makers cannot claim a food is “complete and balanced” unless they can prove this in one of three ways:

• A feeding trial confirms that the food was able to maintain a population of test subjects for a determined period.

• A laboratory analysis confirms the food contains nutrients in amounts established by AAFCO as necessary for either “maintenance,” “growth,” or “gestation and lactation.” If the label claims it meets the AAFCO nutrient profiles for “all life stages,” the food must meet or exceed the requirements of the AAFCO nutrient levels for growth and gestation/lactation.

• A dog food in the same “product family” as a similar, “lead” product that passed a feeding trial may make the same nutritional adequacy statement as the lead product. To qualify as a “family member,” the family and the lead product must be of the same processing type (both are extruded kibble, for example) and contain the same approximate amount of moisture, a similar amount of metabolizable energy (as determined by a metabolic study), and similar amounts and ratios for crude protein, calcium, phosphorus, zinc, lysine, and thiamine (as determined by laboratory analysis).

There are three ways a product can earn its AAFCO statement, but only two different statements found on dog food labels: a “nutrient levels” or a “feeding trial” statement. That’s because AAFCO allows a product that meets the family requirements to use the lead product’s feeding trial claim, even though the family members didn’t pass a feeding trial. To complicate matters, pet food companies don’t (and don’t have to) disclose to consumers which food is the lead product that passed a feeding trial and which foods are the family members.

Feeding trials are considered by most veterinary nutrition experts 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.

AAFCO Feeding Trial Protocols

Of course, to be of comparative value, a feeding trial needs to be standardized, so all products that “passed” a trial can be considered equally adequate. That’s where AAFCO comes in.

Contrary to its official-sounding name, AAFCO is not a regulatory power; it’s an advisory body comprised of state feed control officials (the voting members) as well as representatives from the pet food industry, the Food and Drug Administration (FDA), the United States Department of Agriculture (USDA), and other interested groups.

AAFCO researches and discusses policies that affect the feed industry, and drafts model regulations that feed control officials can bring home to their states for modification and/or adoption. Committees are appointed to study specific topics, such as ingredient definitions, labeling, manufacture, inspection, and enforcement; the committees then bring their analyses to the entire body for review and action. All of the individual American states enact and/or codify most or all of AAFCO’s suggested regulations for the manufacture and sale of animal feeds within their borders.

AAFCO designed the standardized protocols used by pet food makers to test the nutritional adequacy of their products through feeding trials. There are slightly different rules for different trials, depending on the claim the manufacturer wants to make for its product. The possible categories for dogs are:

• Adult maintenance
• Growth (puppy food)
• Gestation/lactation
• All life stages

We’ll examine each of these in detail.

Tried as Adults

For an adult maintenance claim, a minimum of eight healthy adult dogs (at least one year of age) and of optimal body weight are required to start the test. Eight control dogs meeting the same description are required for a control group and are fed a different food. While eight is the minimum number for each group, a minimum of 30 dogs must be monitored to establish values for “colony averages” of weight and blood test results before and after the trial.

Because a minimum of 30 dogs are required to establish the colony averages, the number of dogs used for both the control and the tested food groups are generally more than the minimum of eight. The distribution of different breeds in the two groups should be the same.

The adult maintenance test runs for a minimum of 26 weeks. The test diet is the sole source of nutrition (besides water, which must be provided at all times) for the dogs in the test diet group; the control group is fed a different food, one that has already passed a feeding trial for the same claim.

All the dogs are weighed at the beginning of the test, weekly, and at the end of the test. They also receive a complete physical inspection by a veterinarian before and after the test, and are evaluated as to general health, body, and hair coat condition. After the test, blood is taken from each animal and the results for hemoglobin, packed cell volume, serum alkaline phosphatase, and serum albumin are recorded.

Some dogs, not more than 25 percent of those starting the test, may be removed from the trial for “non-nutritional reasons” (say, one develops cancer or somehow breaks a leg) or for poor food intake (meaning the dog is not eating enough). Dogs can be removed for the latter reason only in the first two weeks of the test. A necropsy must be performed on any dogs that die for any reason during the test and the findings must be recorded.

The test food will fail the trial if any dog on the test diet shows clinical or pathological signs of nutritional deficiency or excess, or if any dog on the test diet loses more than 15 percent of its initial body weight.

The average weight of the entire test group is also taken into account; the food fails if the entire group’s average body weight change (from the start of the test to the end) is more than -10 percent or if it is less than the average of the control group’s body weight (minus an allowance for normal variation).

The average values from the dogs’ blood tests can also cause a food to fail. The test group must have adequate values for hemoglobin, packed cell volume, and albumin and must not exceed specific values for serum alkaline phosphatase.

A pet food manufacturer can employ the protocol for either a growth or a gestation/lactation study – both are considered more rigorous than the adult maintenance protocol – and still make the “adult maintenance” claim if the food passes the trial.

Finally, in case it wasn’t absolutely clear, feeding trials are not terminal; they do not require the euthanasia of any dogs.

Puppies on Trial

The protocols are similar for the growth and gestation/lactation trials. The major differences are as follows:

In order to make a growth claim, a minimum of eight puppies from three different mothers are required to start the test, with a control group of eight puppies from three different mothers. The distribution of pups of different breeds and sex in the two groups should be the same. A total of 30 puppies are required to establish colony averages. The puppies must be weaned, but no older than eight weeks, when the test is started. The test runs for a minimum of 10 weeks.

The post-trial blood test measures the same blood values as the older dogs, except for serum alkaline phosphatase, which is not measured.

The average body weight gain in the pups fed the test diet must be no less than either 75 percent of the colony average, with averages for males and females determined separately, or the colony average minus 2.33 times the standard error.

For gestation/lactation trials, enough females must be used to ensure that a minimum of eight pregnant females start the trial. The females must be in at least their second heat period and be at least one year of age.

There is no size or breed requirement, but the females must have been bred to dogs of the same breed. Breed distribution must be the same in both the test and the control group. A minimum of 30 pregnant females must be used to develop colony averages.

The test begins at or before estrus, and ends when the puppies are four weeks of age. The females are weighed at breeding, weekly during gestation, within 24 hours after whelping, weekly during lactation, and at the end of the test. The puppies’ body weights are measured within 24 hours of birth, weekly, and at the end of the test.

The litter size is recorded at birth, one day of age, and at the end of the test. Stillbirths and congenital abnormalities are recorded. Both the mothers and puppies are given a physical examination by a veterinarian at the end of the test and general health, body, and hair coat condition are recorded.

As with the other trials, the diet fails if any female or puppy shows clinical or pathological signs of nutritional deficiency or excess. Additionally, 80 percent of all the puppies that survived past the age of one day must survive and successfully finish the test.

Also, the females must gain weight during their pregnancies, and the average percent body weight change of the females – from breeding to the end of the test – can’t be too great. The litter size of the puppies in the test group must be at least 80 percent of the colony average, and they must have an adequate average weight (with respect to the colony averages) at the end of the test.

The required blood test results are similar to those required for the growth trial.

All the World’s a Stage

To earn the right to make the claim that a food provides adequate nutrition for all life stages, the product must pass the gestation/lactation trial and then be fed to the puppies produced from the litters whose mothers ate the test product in a growth trial. This is the most stringent test used to confirm a dog food’s nutritional adequacy.

However, while the term “all life stages” suggests a dog can eat the food from puppyhood to grave, this probably shouldn’t be taken quite that literally, since the inclusion of senior dogs is not required in any AAFCO feeding trial protocol.

A consumer might also mistakenly conclude that an all life stages food is guaranteed to benefit dogs with various health problems – dogs with poor kidney function, cardiac conditions, diabetes, etc. Not so. In fact, there are no AAFCO feeding trial protocols that prove a “medical benefits” claim.

Good, Not Perfect

The most significant criticism of the feeding trials method of providing nutritional adequacy – aside from the alleged cruelty of the tests, which we’ll discuss below – is that foods are tested for just six months. Many people feed their dogs the same food for years on end, unaware that there may be no scientific proof that dogs can thrive on the formula for years on end. Six months may not be enough time for the effects of any nutritional excess, deficiency, or imbalance to express itself to the point of detectable health problems in the test dogs.

AAFCO is cognizant of this concern. The serum alkaline phosphatase assay was added to the adult maintenance trial protocol in recent years, in an effort to detect calcium deficiencies that may not otherwise be detectable in a physical examination at the end of the six month trial.

As another example, there have been reports of foods found to contain levels of taurine that proved too low to prevent the development of cardiomyopathy in consumers’ dogs – after passing feeding trials. And the protocols don’t address the unique nutritional needs of breeds that are prone to genetic disorders that require specific nutritional therapy, such as Bedlington Terriers (which require a diet especially low in copper and high in zinc).

Another criticism has to do with the relatively small number of dogs required in the tests. As we detailed above, 16 is the minimum number of dogs needed to pass a feeding trial to prove an adult maintenance claim; this includes the dogs eating the tested food and the “control” group that eats another food. Theoretically, just eight dogs could “prove” the nutritional adequacy of a food that becomes the sole source of nutrition for millions of dogs. In actuality, the studies are generally more populous than the bare minimum required, but the more dogs that are used, the higher the cost of the tests, so it follows that no formula is likely to be tested on thousands of dogs.

It’s also potentially problematic that the lives – and therefore, the nutritional needs – of dogs in labs aren’t terribly similar to most of our dogs. Of course, as you’ll see below, we’d like to see the living conditions of test dogs come to more closely resemble those of our dogs.

Metabolic Studies

These studies have become far more common since AAFCO established a third way (the “family” protocol) for a product to earn its proof of nutritional adequacy; this method requires a metabolic study of each prospective family member.

Metabolic studies require a dog to be kept, for a minimum of five days, in a cage with a slatted metal floor. This floor allows the lab staff to collect every bit of the dogs’ urine and feces for analysis. The total amount of metabolizable energy in the food can then be determined by tracking the dog’s food intake to calculate the gross energy consumed and subtracting the energy in the waste products.

Life in a Lab

Now for the bad news. PETA is right about one thing: Life in a cage sucks for any animal. This is not quite the same thing as saying life in a cage is inherently cruel. But dogs are pack animals, and a solitary existence, devoid of physical contact with other dogs or other enrichment or socialization does not meet what we would consider a required minimum mental health standard.

And, while any individual feeding trial is just six months, and metabolic studies are just five days in duration, the “time served” by most lab dogs is at least several times those numbers. This is because labs tend to use the same animals over and over again in back-to-back trials. Lab directors defend this recycling, saying it’s surely better to reuse a small population of dogs for a lifetime of trials than to subject a larger number of dogs to just one or a few of the tests.

The good news is that dogs in feeding trials don’t have to live in cages. They can be kept in runs that have access to fresh air and time outdoors, in compatible groups or pairs, and have opportunities to participate in social playtime. Of course, as any kennel operator knows, this sort of facility is far more expensive to staff and maintain. Only pet food makers that really care about its test subjects (or at least, the opinions of consumers) will allocate the extra money for their test dogs to live in an adequately enriched and pleasant environment.

Under pressure from consumers – which can be at least partly attributed to PETA’s media campaign – both the pet food companies and the contract labs they sometimes use are beginning to discuss and consider improving the living conditions for the dogs in their feeding trials.

There have been a number of articles in Pet Food Industry about the industry’s need to improve the welfare of its test subjects. Several pet food industry conventions have presented speakers who made recommendations for lab animal socialization and enrichment programs. And most major pet food makers have published (or made available upon request) a description of their test animal welfare programs and goals.

Reform Still Needed

Unfortunately, this is not to suggest that everything is fine, now. Even the companies that advertise that they provide the very best living quarters, enrichment programs, and provisions for retirement or adoption of their old test dogs don’t invite confirmation of their claims.

For example, one contract lab, Summit Ridge Farms in Pennsylvania, has taken out ads and sent press releases to pet food industry publications, announcing the construction of a huge “puppy playground” and other innovations constituting “the beginning of a long-term environmental enrichment expansion” at its facility. A press release published in Pet Food Industry quoted Mike Panasevich, president of the company, as saying, “We are extremely happy and proud of our facility and the enrichment programs currently in place.”

However, this same executive wouldn’t return our repeated calls or e-mail messages to discuss these positive developments with us – not what you’d expect from someone who had genuine innovations to show off.

This opaqueness makes it appear that the industry still has a lot to hide. And, unfortunately, PETA’s infiltration of a contract lab has now afforded the entire industry with a convenient excuse to hide its work, citing security risks.

Throwing Down a Gauntlet

It seems to us that it would be in the best interests of pet food companies and contract labs to keep their test subjects as happy and comfortable as possible. We imagine this would help the foods produce and maintain health in the test dogs, so the products pass with flying colors.

But the companies would also be wise to do the right thing by their test animals as a marketing technique. For our part, we’d be thrilled to promote a pet food made by a company that could and would prove that its labs provide the best possible living conditions for its test subjects, with comfortable living quarters, plenty of socialization and exercise, and attentive veterinary and behavioral care.

One would think that a policy of openness and full disclosure at a facility that was truly doing everything right for the test dogs and cats would be a fantastic selling point for a pet food maker. But we haven’t found a company that’s willing (or able?) to take that challenge.

-Nancy Kerns is WDJ’s editor.