Proper Raw Diets and AAFCO
(Pet Food Forum 2002)

Raw diets – homemade and commercial -- are probably the fastest growing segment of the dog food market. They are here to stay.

In the last two years, we also saw a continuation of error riddled anti-raw articles published in prestigious dog and veterinary journals. These articles, specifically one in the JAVMA, have unfortunately led to a loss of credibility of petfood science among the growing "alternative" or "holistic" market.

Many of the alternative or holistic pet owners don’t know how important AAFCO has been to the improvement in pet foods. Many new raw diet manufacturers reject all "establishment" petfood science and do not understand that AAFCO nutrient profiles represent decades of the best science available.

The petfood industry needs to discuss how to regulate proper raw diets. Raw diets, because they are raw and usually variety-based, require slight modifications to AAFCO protocols and regulations. Proper regulation is as important to my market as it is to all petfood markets. We propose for AAFCO to:

1. regulate raw diets just like dry and canned

plus:

2. require safe handling instructions

3. modify AAFCO protocol to accommodate variety based diets

And, for those considering studies of raw diets, I ask, please do it right. We have a wealth of data that can help you. We have much to discuss together, poorly researched anti-raw articles just push people further apart.

In order to facilitate discussion, I am going to update you on the market for raw diets. Then Spencer will discuss AAFCO protocol and variety-based raw diets. We’ll then conclude and have ample time for discussion.

First, I will define what we mean by proper raw diets.

Let’s emphasize proper. Quite frankly, the best looking dogs I’ve ever seen are on proper raw diets; the worst looking dogs are on improper raw diets.

I define proper raw diets as "complete and balanced foods or feeding programs, using human-edible ingredients, and processed and handled according to human-edible standards. Proper commercial products have clearly marked safe handling instructions and AAFCO statements".

Broadly speaking, there are feeding programs from Schultz, Volhard, and Billinghurst (who popularized Bones and Raw Food or BARF) where the dog owners prepare the food themselves; and complete and balanced commercially available foods meeting AAFCO guidelines, like Steve’s Real Food® for Pets and others. Spencer will elaborate later.

Most raw diets are based upon the natural diet of dogs. Think of the dog eating a mouse or rabbit. The dog would eat raw meat and organs, crushed bones and the prey’s digestive system, which would be full of finely crushed vegetables, fruits, and enzymes. The natural diet contains very little, if any, grain.

The dog would eat a wide variety of foods. My market and I believe it is not wise to feed a dog, a cat, or a human, the same food week after week, month after month. We recommend that our customers feed our chicken food one day, our beef the next, then turkey and so forth. Hence the need for modifiying AAFCO protocol for variety based diets.

Proper raw diets are expensive: Costs range from equal to the cost of super-premium kibble foods for good shoppers who make their own food, to three times the cost of super-premium foods for some of the commercial raw meat-based diet brands.

Now that we’ve seen what raw diets are, let’s take a closer look at the commercial products, the products that are of the most interest to AAFCO.

We’re going to look at who sells, who buys, how they use the food, and where raw diets are most popular. I will use data from my company and assume it is indicative of the raw diet industry as a whole.

Our retailer market consists of, independent petfood stores, natural food stores, veterinarians, groomers and kennels. Some raw diet manufacturers sell direct to dog owners and buying groups, we do not.

Today, with premium petfoods available in grocery and mass-merchandise, independent petfood stores need differentiated, high-margin products like proper raw diets in order to survive. They certainly can’t compete with mass on price. About 450 independent petfood stores have invested in freezers and now sell my products. Overall, I think about 8% of all independent petfood stores sell raw products.

I recently visited about 50 independent or "Mom and Pop" stores in Texas and New England. Retail can be a tough life, the owners sometimes work very long hours. For the most part, though, they are not knowledgeable about AAFCO.

Three of our top ten retailers are veterinarians. We work with holistic and a growing number of traditional veterinarians. There are hundreds of veterinarians nationwide who are selling or recommending properly prepared raw meat-based diets. I think that number is tripling every year.

Natural food store shoppers understand my products. About 40 natural food stores now sell my products. Our success in these stores tells us that Steve’s Real Food will dominate dog food sales in these stores. This market seems to be extremely brand loyal.

Manufacturer sells direct

To breeders, dog clubs, buying groups

Who uses

The more Starbucks in a location, the more customers we have. The West, increasing growth in New England and Texas.

The future

I predict that the major categories will be dry, canned, and raw, with just as many dogs eating raw as canned. There will be a thriving community of independent pet food stores specializing in selling proper raw diets.

Some raw diet manufacturers will grow by selling their products direct to dog buying groups, bypassing the traditional distributor to retailer channels completely.

And, in the next 20 years, I expect Steve’s Real Food and other raw diet companies to have sponsored peer-reviewed science on raw diets. That’s what Spencer will speak about.

AAFCO Methods and Raw Diets

My section will cover three general topics:

1. I will review feeding methods for some of the more popular raw feeding methods
2. I will describe current AAFCO feeding protocols and the rationale behind them, and
3. I will enumerate ways in which raw diets may be examined using current or modified protocols.

I’ll conclude with an overview of some of the avenues of research to help examine and qualify raw diets.

Let’s look at a few of the more popular methods of raw feeding. This will help us understand how they may be evaluated using a current or modified AAFCO feeding protocol.

Raw feeding began as a do-it-yourself, homemade regimen. Pet owners took it upon themselves to procure raw meaty bones, vegetables, and supplements and chopped, blended, and mixed their own meals for their pets.

There are two main types of raw diets:

1. Those that feed the same ration at every meal. These are most analogous to traditional diets.
2. Those that feed different rations at each meal. This more closely reflects the practices of most DIY-feeders. Generally, these diets are fed in periodic cycles.

Since we’re more familiar with the first regimen, let’s examine the variety-based rations first. Also, keep in mind that many of the commercial raw diets out there can be thought of as simplified versions of variety-based diets.

There are many competing versions of the homemade raw diet out there, popularized primarily by the Internet, word-of-mouth, and self-published texts. Differences between diets generally consist of ingredients, ingredient sources, feeding practices, and even ideological differences. For example, some regimens differ sharply over whether to feed "sugary" vegetables such as carrots or beets.

In general, though, homemade diets have several features in common:

1. All, of course, are fed raw and use human-grade ingredients.
2. All usually feed a ratio of about 60% RMB’s and 40% fruits/vegetables and supplements.
3. With few exceptions, no grains.

Let’s investigate these aspects. Most sources, perhaps intentionally, do not provide a rigid definition of a Raw Meaty Bone. Generally, chicken wings, chicken backs, and so on are the most commonly used, though beef knuckles, turkey necks, and other cuts are not unheard of. The idea is to provide a source both of meat AND of calcium/phosphorous, as well as to give the dog something to chew on, for dental health.

Next, these diets generally have about 60"% RMB’s and organ meat, with the other 40% made up of fruits, vegetables, the occasional dairy product, and supplements. Most owners find that they have to mix the vegetables in with the RMB’s using a grinder; otherwise, dogs can ignore the vegetables.

Notably, most of these diets contain little to no grain. This stems, like most aspects of raw diets, primarily from the observation that the diet of feral carnivores does not contain grain or other significant sources of carbohydrates beyond what is consumed from the stomach contents of their prey.

As you may imagine, these methods require quite a bit of work. Until recently, many do-it-yourselfers made do with sourcing ingredients on their own from local butchers and markets. Finding butchers willing to grind raw bones has always been a challenge.

Small companies are beginning to market products based on this approach. These take the form of frozen eat/vegetable mixes, vacuum-packed RMB’s, and so on.

The basic assumption behind these diets is that while not every meal in and of itself meets canine nutritional requirements, the diet as a whole does. While there may not be, at this time, much commercial interest in these diets, it is obviously of interest to veterinarians and animal nutritionists (not to mention their human and canine clients) that the nutritional quality of these diets be assessed in an objective manner. It will likely turn out that some homemade raw regimens are balanced, and some not.

Let’s move on to a discussion of complete and balanced raw diets, like Steve’s Real Food. These, as I mentioned, are more analogous to traditional canned or kibbled rations. Each meal meets all nutrition needs for a healthy dog at the appropriate life stage. Proper raw diets, as Steve mentioned, meet the AAFCO nutrient requirements for the appropriate life stage. Most prepackaged raw diets are frozen diets, however, an increasing number have either freeze-dried versions or are exclusively freeze-dried. We recommend, as Steve noted, rotating varieties on a periodic basis. Many of our customers mix our food with traditional kibbled diets.

AAFCO Protocols

1. By meeting the Association of American Feed Control Officials’ published nutrient profiles for the appropriate life stage

Or

1. By performing feeding trials as prescribed by AAFCO protocol.

I’ll discuss both, with an emphasis on feeding trials.

Historically, a difficulty in establishing nutrient profiles had been a lack of an agreed-upon standard for assessing nutritional adequacy. The NRC formulated guidelines for canine and feline nutrition in the 1970’s, but these were based on purified diets and did not take factors like bioavailability and ingredient quality into account. The present AAFCO guidelines were formulated to address these issues in a real-world context.

If a food meets these analysis requirements, even on paper, it may use the phrase "complete and balanced" on its label. There are slightly different minima for growth & reproduction and adult maintenance. A food that meets the growth profile may advertise itself as a food fit for all life stages.

There are several reasons why nutrient profiles alone do not provide a complete picture of nutritional adequacy.

First, they do not completely address issues of nutrient bioavailability and interactions. Transition metal binding to grain phytates has been a perennial problem both in human and veterinary nutrition. AAFCO profiles do, in fact, disallow copper and iron oxides due to their poor bioavailability. The NRC, I believe, is beginning to mobilize a much-needed investigation into nutrient bioavailiability in pet diets.

Studies have, in fact, confirmed malnutrition in foods that meet AAFCO nutrient profiles. Huber et al reported clinical symptoms of zinc and copper deficiencies in a regionally marketed ration that met the nutrient minima.

Third, it is possible for a manufacturer to fulfill the letter of the regulations by simply using a spreadsheet program to analyze their diet, without chemical substantiation. Nothing in the AAFCO Official Publication requires a manufacturer to perform wet chemistry on their diet beyond a simple guaranteed analysis of crude protein, fat, and so on. This technique requires a small degree of mathematical and software sophistication but is easily within the range of a small manufacturer.

Hence, the preferred method of nutritional substantiation has become the AAFCO feeding protocols. Let’s look at some of the major aspects of feeding trials so that we can understand how they can be used to determine the nutritional adequacy of raw diets.

The advantage of feeding trials is simple. They give a better overall picture of nutrient bioavailability. By tracking a dog through several months of food consumption, nutritional deficiencies can become clinically obvious.

AAFCO-sanctioned feeding protocols look at three main areas:

1. Clinical signs of malnutrition
2. Blood parameters like hemoglobin and so on.
3. Weight change, as appropriate for the life stage in question

Scientific control is established by using one of two methods. The trial administrators may compare the test group against a concurrent control group. Alternately, they may make comparisons with the historical averages (in blood parameters, weight gain, and so on) of the established colony. The diet fed to either the control group or the colony must itself have passed an AAFCO-sanctioned trial for the appropriate life stage.

These criteria change from one life stage to another, so let’s look at the first task at hand: substantiating a raw diet for all life stages. This will require the regimen to pass both a gestation/lactation AND a growth protocol.

The gestation/lactation protocol requires that at least eight pregnant bitches begin the test. All must undergo a physical examination by a vet. Notably, no breed is specified, though the stud must be of the same breed as the bitch. Also of note is the requirement that the same diet be fed throughout the trial. As we will see, this creates problems with the variety-based raw diets.

To add, a minimum of eight puppies from at least three different bitches is required for growth trials.

Blood parameters are also used to assess nutrition on a sub-clinical basis. Analogously, blood parameters are evaluated with respect to historical colony averages or concurrent control group values. Hemoglobin is generally used to track transition metal nutrients (iron, copper, and so on). Packed Cell Volume measures the volume of blood occupied by red blood cells, and is used principally to detect anemia caused by one of many nutrient deficiencies. Albumin analysis can reveal deficiencies of protein and other nutrients. Adult maintenance trials add a serum alkaline phosphatase assay in an effort to detect calcium deficiencies that don’t evidence themselves in a non-growth/reproduction life stage.

Trials have varying lengths. Gestation trials, for example, begin at estrus, and end when the puppies are 4 weeks old. Adult maintanance trials run for a prescribed 26 weeks. In all cases, the test diet can be fed ad libitum, or according to energy needs.

Food consumption is recorded on a daily basis. Bitch weight is recorded on a weekly basis during gestation and lactation, as well as at breeding, at the end of the test, and within 24 hours of whelping. Puppy weight is measured within a day of whelping and weekly thereafter. Litter sizes are noted.

Puppies are given physical examinations within 72 hours of birth and at the end of the test. Up to 25% of bitches may be removed from the test for non-nutritional reasons (like poor food intake, for example), but only in the first two weeks of the trial.

Tests are interpreted in a relatively straightforward manner. If any puppy or bitch shows clinical signs of malnutrition, the trial fails. Eighty percent of all puppies that survive 24 hours postpartum must finish the test. The average weight gain during gestation must compare favorably to either the control group or the historical colony average. Here’s an overview of the statistical interpretation:

If the diet is being compared against the colony’s historical average, the weight gain may not be less than the historical average minus 2.33 times the standard error. Against a concurrent control group, the average weight gain may not be less than the average for the control group minus an allowance for normal variation. Somewhat more complicated than a simple standard error calculation, this involves pooling the estimate of the standard error of both the test and control groups.

Similarly, the average weight gain of the puppies, postpartum, must not be less than either 75% of the historical average, 2.33 times the standard error of the historical average, or the control average minus the variance allowance. The litter size, also, is compared against the control group or the historical averages.

To be sure, problems do exist with feeding trials as they currently stand. There is a difficulty in establishing historical colony parameters, especially when the colony has been fed a qualitatively different diet than that used for the test. While protocols do specify, for example, that fresh water be provided to all animals, PCV counts may be altered when comparing a dry, kibbled ration to historical figures based on a wet, canned regimen due to fluid and electrolyte loss.

More relevantly, AAFCO protocols do not address long-term nutritional deficiencies. As I mentioned earlier, serum alkaline phosphatase assays were added to adult maintenance trials in an effort to detect calcium deficiencies in adult maintenance trials. As reported recently, it appears that at least one commercially available dog food, having passed feeding trials, does not provide enough dietary taurine (or taurine precursors like methionine and cysteine) to ward off cardiomyopathy.

Also, feeding trials need not address breed-specific disorders like the copper metabolism of Bedlington Terriers or large-breed skeletal issues. All that is required is that the breed distribution of the control (or colony) and the test group be similar. Beagles are commonly used as feeding trial participants. It can be argued that a diet that supports a beagle puppy through to adulthood may not be the most optimal for a Great Dane puppy, or vice versa. Indeed, recent aspects of taurine supplementation controversy seem to indicate that cocker spaniels may lack sufficient quantities of taurine-synthesizing enzymes.

Evaluating Raw Diets with Feeding Trials

Little information exists as to the expected weight gain or loss of a dog or puppy upon the introduction of a raw diet. It is assumed, however, that it will be similar to the gain on a conventional diet. Anecdotally, dogs on proper raw diets tend to find their ideal weight over the long term, though initial weight loss is common. This is probably due to the relative lack of carbohydrates in raw diets.

The wording of AAFCO protocols does not allow for variable feeding regimens. In fact, it specifically disallows them with the phrase:

"The same formulation shall be used throughout the test…"⁶

This excludes an evaluation of variety-based raw diets. Still, since AAFCO feeding protocols are, currently, the most widely accepted means of assessing nutritional adequacy, a modified protocol may be the best way to determine the completeness of homemade diets in a non-commercial setting.

It is suggested, therefore, that an evaluation of such diets be subject to the same clinical parameters of the AAFCO protocol, but be extended to cover at least three or four iterations of the diet cycle. Should AAFCO desire to make a permanent category for variety-based diets, it is urged that the current protocols be amended to include variant rations.

Since many raw diets rotate their ingredients on a periodic basis, the length of protocols could be modified to assess a complete number of cycles. For example, adult maintenance trials could accommodate 6 to 8 4 week cycles, or growth trials 2 or 3.

So, to summarize, I’ve given a general overview of popular raw feeding methods, I’ve described current AAFCO-sanctioned methods of nutritional substantiation, and I’ve talked a little about how to properly evaluate these diets using substantiated, peer-reviewed methods. The next few years should be very interesting for both the proper raw market and the development of the science behind it. Thank you.