Last July, the FDA’s Center for Veterinary Medicine announced that it had begun investigating a “potential link” between “certain pet food ingredients” (peas, lentils, potatoes, and other non-grain sources of starch often found in “grain-free” pet foods) and the rare but deadly canine heart disease dilated cardiomyopathy (DCM).
The FDA’s investigation was prompted by a group of American veterinary specialists from major research universities, including the University of California at Davis and Tufts University. Claiming to have noticed a “recent spike” in DCM cases, several of these same veterinarians also banded together to co-author an article that appeared in the December 2018 edition of JAVMA, the most widely-read veterinary medical journal on the planet. Their article, purporting to be a review of the current evidence surrounding diet and DCM, was entitled “Diet-Linked Dilated Cardiomyopathy: What Do We Know?”
Together, the JAVMA article and the FDA’s announcement helped to catapult diet-associated canine DCM from the fringes of the veterinary nutritional science community into the pet-owning mainstream. The New York Times covered the story under the headline “Popular Grain-Free Dog Foods May Be Linked to Heart Disease.” Slate went even further, publishing a news analysis with the alarming headline “Please Stop Buying Your Pets Grain-Free Foods.”
Nearly half of the pet food sold in America today is grain-free. And the suggestion that products consumed by tens of millions of beloved family pets could cause a deadly disease sparked panic and outrage across the country. Indeed, over the past year interest in DCM has far outpaced the disease itself. Although fewer than 600 cases of canine DCM have been reported to the FDA in connection with its investigation, a newly-formed Facebook group called “Taurine-Deficient (Nutritional) Dilated Cardiomyopathy” swelled to more than 60,000 members in just a few months.
The veterinary community has been just as keenly interested as the pet-owning lay public. In the eight months since its publication, the JAVMA article has already been downloaded more than three times as often as any other article published in the prestigious journal over the past year. At more than 80,000 downloads as of July 2019, it may in fact be the most widely-read veterinary science article ever written.
But that might be about to change. Because last Friday I delivered a letter to JAVMA’s editorial board demanding that the article be retracted immediately. (Here’s a link to the retraction materials so you can read them for yourself.)
My letter (and the 40+ pages of evidence and other supporting materials) is the culmination of an investigation that lasted nearly a year and involved a diverse range of methods. I interviewed several of the JAVMA article’s authors along with a host of other veterinary nutritionists and animal biosciences professors, commissioned an independent laboratory to perform key biochemical analyses ignored by the authors, conducted statistical analyses that reveal irregularities in the data used to support the article, and pursued public records investigations at both the state and federal level in order to get my hands on e-mails and financial records that had not previously been disclosed to the public.
I’m biased but it seems to me that, by any fair reading, the case for retraction is a slam-dunk. (Again, you can read the complete retraction package and offer your feedback here.) There are professional organizations that set ethical standards concerning such matters, and the present case clearly meets them . To wit, the retraction package was co-signed by more than 200 veterinarians, animal scientists, and other stakeholders. And one of the most well-regarded and visible animal nutrition PhDs in the world echoed the call for retraction just a few days later.
Whether JAVMA decides to do the right thing and retract the offending article is another matter altogether. (If you want to tell the editors how you feel about the matter, you can do so here.)
Either way, the retraction demands are just one part of my broader, ongoing effort to correct the scientific record surrounding this issue. I have also outlined the glaring methodological problems and suspicious data irregularities in recent DCM studies conducted by two of the authors of the JAVMA article and filed a federal lawsuit against the FDA. As part of my investigation, I asked the FDA to disclose records concerning its relationship with the JAVMA article’s authors as well as three pet food companies to which several of the authors have financial ties. For some reason, the FDA has stonewalled these efforts and has refused to produce any information whatsoever concerning its investigation. My lawsuit is an attempt to force the agency to comply with its duties under federal transparency laws.
I have been motivated to do all this for several reasons. First of all, as the author of a book about bad science and financial conflicts of interest in the veterinary nutrition community, I’m one of very few key opinion leaders in a position to notice the problems and actually speak out about them. Secondly, as the founder and CEO of a “BEG” pet food company that most definitely does not give our customers DCM, I have a direct motivation to stop what amounts to the defamation of my business. Lastly, as pet-owner myself, it is hugely important to me that the most prestigious veterinary journal in the world not be allowed to become a source of propaganda.
The retraction demand package is densely packed and nearly 50 pages long. It’s not light reading. If you don’t have time to digest the whole thing, here are the ten most important take-home conclusions:
1) There is no evidence that canine DCM is associated with “BEG diets,” “grain-free” diets, or any specific ingredients. To be clear, I’m not just saying there isn’t ironclad evidence that these diets cause DCM. I’m saying there isn’t even evidence that they are correlated with the disease. Not one single study. There are seventeen places in the JAVMA article where the authors state that “BEG diets” are “associated” with DCM. That statement is demonstrably false.
2) There is no evidence that canine DCM rates have increased in recent years. Again, none whatsoever. Not a single study. The idea that DCM diagnosis rates are “spiking” is based entirely on the claims of a handful of veterinarians with a direct financial motivation to promote it.
3) DCM is really, really rare. DCM is a real and serious disease, but it’s also incredibly rare. So rare in fact that you’re far more likely to be struck by lightning than to have a dog diagnosed with it, regardless of what the animal eats. The best data currently available says that fewer than 1 in 100,000 dogs will be diagnosed with diet-associated DCM. The odds of being killed by an asteroid are slimmer.
4) The three veterinarians at the heart of the DCM investigation — Dr. Lisa Freeman, Dr. Josh Stern, and Dr. Darcy Adin — all have financial ties to one or more of Hill’s Pet Nutrition, Mars Petcare, and Nestle-Purina Pet Care. Because these companies focus heavily on grain-containing pet foods, they stand to gain phenomenally if the market for “BEG diets” drops out. And the three vets who both instigated the FDA’s investigation and wrote the primary academic articles about DCM all have financial ties to one or more of them. What a coincidence!
5) The leading academic article about DCM wasn’t peer-reviewed. The article was published in JAVMA, the most widely-read peer-reviewed veterinary journal on the planet. It has been hugely influential in the veterinary community (it’s been downloaded more than three times as often as any other JAVMA article published in the past twelve months). But its authors (the three vets mentioned above plus two others) cleverly avoided the peer-review process by mislabeling the article an op-ed, the one piece of writing that JAVMA doesn’t peer-review. In reality, the article isn’t an op-ed, it’s a factually-focused evidence review. (The article is entitled “Diet-Associated Dilated Cardiomyopathy in Dogs: What Do We Know?”) Because the article wasn’t peer-reviewed, it contains all sorts of false statements about DCM, the most glaring of which being that there’s evidence that the disease is “associated with” specific pet foods. Again, there’s literally no evidence of this whatsoever.
6) There have been only two diet-associated DCM studies published in recent years, and they both contain big methodological problems. The first contains a very suspicious data irregularity. This study involved separating all DCM cases from a specific veterinary hospital into two categories: those on grain-free diets and those on grain-containing diets. At a preliminary stage of the research, the authors (the aforementioned Dr. Adin and some colleagues) reported that 22 of the DCM-positive dogs in their study group were on grain-free diets while 27 were on grain-containing diets. (I.e., they found that the majority of DCM-positive dogs were eating grains.) But by the time their final paper was published, the number of DCM-positive dogs on grain-free diets had grown to 36 while the number of DCM-positive dogs on grain-containing diets had plummeted to 12. For some reason, fifteen of the DCM-positive dogs on grain-containing diets just vanished from the study group, making grain-free diets appear much more common among DCM cases. In the retraction materials I explain why the odds of this outcome occurring due to random chance are less than 1 in 10,000.
7) The other recent diet-associated DCM study also contains a suspicious methodological abnormality. The veterinary community already knows of one sure-fire dietary cause of DCM: feeding a diet that doesn’t contain enough of the amino acids cysteine and/or methionine. There’s abundant evidence that these types of deficient diets can cause DCM and it’s described in every leading veterinary nutrition textbook on the planet. But, for some reason, Dr. Stern and his colleagues just ignored the issue in connection with their recent DCM study. Which is particularly strange because the research team is associated with one of the leading amino acids laboratories in the country, the one at UC Davis. Well, I went back and actually did this work (I sent the diets in the study off for independent amino acid analysis) and I found that 60% to 80% of the DCM-positive dogs in their study in fact weren’t consuming their recommended daily allowance of cysteine and/or methionine. (You can read the reports for yourself here.) In other words, a well-understood dietary cause of DCM was at play in the majority of dogs in their study. And it has nothing to do with grains. And for some reason the authors just completely ignored it.
8) Dr. Freeman has used this same playbook before to great effect. Prior to leading the charge against “BEG diets,” Dr. Freeman was already one of the country’s leading proponents of the notion that raw-ingredient pet foods (another group of products that threaten Nestle-Purina, Mars, and Hill’s) are a great public health danger and should be avoided by pet owners at all costs. Even today her 2013 JAVMA paper on the subject is still one of the journal’s most-downloaded articles. Pick up a phone, call your local veterinarian, and ask her what she thinks about raw-ingredients diets. Odds are she’ll tell you that they are likely very dangerous and have no known benefit. Sound familiar? If you want to understand why Freeman’s opinion about raw diets is misleading and incomplete, I explain the details here.
9) The numbers just don’t add up. Industry analyses suggest that grain-free pet foods make up about 44% of the overall pet food market. That means they’re being eaten by more than 35 million dogs every day. Include “boutique” and “exotic-ingredient” foods as well and the number rises to well above 45 million. Now, let’s just assume for the sake of argument that every single DCM case reported to the FDA was also eating a “BEG diet.” That’s not the case, but let’s pretend for a second that it is. Even in that case, the vast majority of dogs on “BEG diets” would still be doing just fine. How vast a majority? Not 51%. Not 60%. Not even 70% or 80%. Even in that unrealistic case, some 99.999% of dogs on BEG diets still wouldn’t be getting DCM. Any notion that these kinds of numbers could conceivably represent a causal relationship is facially absurd.
10) Dogs outside of the United States don’t have a problem. “Grain-free” and “exotic ingredient” pet foods aren’t exclusive to the United States. But the DCM “epidemic” is. For some reason, no other country in the world has launched a regulatory investigation of diet-associated DCM. If there was valid evidence that some popular class of pet food products was in fact causing this disease, they surely would have. Pretty strange, right?
Pets in the United States are staggeringly unhealthy. Obesity is deadlier for a dog than a lifetime of smoking is for a human being and more than half of the dogs and cats in the U.S. are overweight. More than 1/3 of all dogs will get cancer in their lifetimes. Diabetes, osteoarthritis — -the list of chronic disease epidemics with links to nutritional factors goes on.
Which is why rejecting the “fake news” about DCM is so important. This most definitely is not a victimless crime. Because if the FDA and the veterinary nutrition community is spending its time, attention, and resources promoting made-up problems, then it’s ignoring the real ones.
Pet owners rely on their veterinarians and government regulators to be fair, honest, and well-informed about the relevant scientific record. Spending time and money trying to solve made-up problems instead of real ones is a shocking abuse of that trust. We deserve better. And so do our pets.