Feeding your dog in the age of the DCM scare

Dilated cardiomyopathy (DCM) is a condition that prevents the heart from pumping properly.[1] It is more prevalent in some breeds, but there have been recent reports of increased rates of DCM in breeds prone to the condition when fed grain-free diets. While this should be investigated further, it is a loose correlation and has been blown out or proportion to the point of terrifying puppy parents that they are feeding their dogs a harmful diet. This condition has been falsely inflated by blog posts, poor or incomplete research, and some major companies taking advantage of the public panic.

So let’s take a quick look at what the danger is.

Rate of occurrence

Reports of about 300 cases of nutritionally mediated DCM (NM-DCM) have been received by FDA. This is horrible for those dogs and families affected, but it’s important to look at the bigger picture as well and realize that given that there are 90 million dogs in the United States alone, this is a very, very low rate of occurrence.

Contrast that with the estimate 60% rate of cancer in breeds like Golden Retrievers. https://www.cbsnews.com/news/why-do-so-many-golden-retrievers-get-cancer-massive-study-hopes-to-find-out/

Since NM-DCM is associated with grain free diets, let’s factor in that as well. Grain free diets are fed to about 37% of the dog population in the US, based on market share[2], so that would give us about 33 million dogs eating a grain-free diet. Again, FDA has reports of 300 confirmed cases. That gives us a rate of occurrence of 0.0009 percent.

Let’s take the worst-case devil’s advocate approach and assume NM-DCM is vastly under-reported by a huge factor, say of 100 times. Our hypothetical situation would mean that only 1 in 100 cases has been reported for a high profile condition and would give us the assumption that there are 30,000 cases instead of the reported 300. With 33 million dogs in the population eating grain free food, that would give us a hypothetical rate of occurrence of 0.09 percent.

To see to an actual rate of 3 percent in the population, we would need to see almost a million dogs eating grain free food be affected with NM-DCM, which is so far from what has actually been reported as to be an absurd expectation. Contrast that with rates of hip dysplasia, pancreatitis, digestive issues, and other conditions that can be more soundly related to diet and have rates of occurrence of double digits in various canine populations.

Information quality

It is also important to look at the quality of the information you are using. Science accumulates incrementally, with early studies providing interesting points, but not being what is considered among scientists to be actionable evidence.

Typically, scientists and/or medical practitioners will notice a phenomenon and create a case study. The case study is a report that explains what someone observed in one or several patients in a practice. At times, case studies will attract further attention from scientists, so the next step often will be for someone to perform a retrospective study.

A retrospective study looks at medical histories from the past. It looks backwards from an outcome that is already decided at the beginning of the study—meaning it assumes a conclusion and then looks at past records to see if those records support the conclusion. This type of study is known for bias and confounders.

Retrospective studies typically are case-controlled, meaning they are prone to selection bias by their very nature. They are typically considered by scientists to help point out points of interest that may need to be studied further. They are NOT considered good scientific proof; rather retrospective studies typically indicate a correlation that needs further research.

Correlation is not causation. The confusion of correlation with causation is quite possibly the largest mistake lay people make when trying to understand the current state of knowledge of a subject. This correlation from retrospective studies is the majority of the data we have on DCM in dogs. It shows us that there's something we may want to investigate further, but that's it. There's no strong proof of cause, and the fact remains that there can be more than one cause. We just don't know yet.

The next step up from a retrospective study is a prospective study. A well-designed prospective study will use an observation to create a hypothesis, then tests that hypothesis using well-controlled variables. Prospective studies give us true incidence rates, relative risk, and more reliable data. Although this is a higher quality type of study, one prospective study, no matter how well done, is not considered a solid foundation on which to make informed scientific claims. It takes many prospective studies to make sound scientific conclusions. We are not even close to this point yet with canine NM-DCM.

We do want to continue investigating this topic. We do want to find the true cause so we can help prevent DCM when possible. We do not, however, want to make rash decisions and potentially harmful changes based on fear rather than fact. We are very emotional about and attached to our dogs. This is not a bad thing. But we need to make our important decisions about their health from a point of objectivity and careful consideration. Knee-jerk reactions are too likely to come back and bite us on our proverbial tails.

How to decide

Please do not make a nutritionally impactful decision about your dog’s diet based on a very low probability of occurrence and on data from lower-quality research, such as case studies and retrospective studies.[*] Because when we do things like this, the law of unintended consequences kicks in and I’m afraid of what we might see as a result.

For those who are interested in this topic and prefer to take a deeper dive, we recommend starting with the Mansilla[3] and Freeman[4] papers. We are excerpting what we feel to be the most important take-home message from the Freeman paper here (emphasis mine).

Preliminary results of a study performed by one of the authors (DBA) found that dogs with DCM that had been eating grain-free diets had more advanced cardiomyopathic changes than did dogs with DCM that had been eating grain-based diets. Unreported results of the study indicated that a subset of dogs clinically and echocardiographically improved after a diet change. Notably, however, some dogs improved after a diet change from one grain-free diet to another, and this finding, along with the differences identified between dogs fed various BEG diets, suggested that DCM was not necessarily tied to the grain-free status of the diet. Taurine supplementation was prescribed for many of these dogs despite the lack of apparent deficiency, and it is unclear what role taurine may have played in their recovery.
Although DCM in some dogs without any apparent taurine deficiency appears to be reversible with a change in diet, with or without taurine supplementation, no cause has thus far been identified for non-taurine deficiency–associated DCM. Possible causes that are being investigated include absolute deficiencies of other nutrients, altered bioavailability of certain nutrients because of nutrient-nutrient interactions, and the inadvertent inclusion of toxic ingredients.
For example, BEG diets could possibly be more likely to have deficiencies of nutrients other than taurine, such as choline, copper, l-carnitine, magnesium, thiamine, or vitamin E and selenium, that have been associated with cardiomyopathies. Although pet foods are required to be nutritionally complete and balanced (unless they have a label statement that they are for intermittent or supplemental use only), that does not always provide a guarantee, and deficiencies could occur if diets do not contain appropriate amounts of all dietary nutrients. Further, a deficiency may occur even if a diet contains the required minimum amount of a nutrient because of reduced bioavailability or interaction with other ingredients in the diet. This may be a concern for diets based on exotic ingredients, whose nutritional properties may not be as well studied.
Researchers are also exploring whether diet-associated DCM in dogs without taurine deficiency may be related to inclusion of a cardiotoxic ingredient in the diet. This could be an adulterated ingredient, as with ingredients containing melamine–cyanuric acid that affected pet foods in 2007, resulting in extensive recalls; a heavy metal; a chemical sprayed on 1 of the ingredients; or even a natural chemical compound in 1 of the ingredients that has toxic effects when fed in large amounts.
Of course, the cause may be even more complicated, such as an interaction between gut microbiota and a dietary factor (eg, trimethylamine N-oxide). It is encouraging that some recovery of cardiac function has been observed in some dogs following a change in diet, with or without taurine supplementation. However, research is needed to identify the underlying cause.

The first paragraph of the above Freeman excerpt is also a good example of how science progresses from a correlation to more solid data. One of the authors initially observed a correlation between grain free diets and DCM. Later, however, this same investigator discovered that the correlation was not sound and is looking further into other potential causes.

My take-home thoughts

In general I do not avoid peas or other legumes as long as they are not in the top few ingredients. I prefer real meat as a first ingredient. And I recommend food rotation to provide nutritional variety.

Remember, feed the dog in front of you, not some vague notion of avoidance of an uncommon condition that likely has nothing to do with the dog you are feeding today.

[*] The obvious exception to this is if you have a dog in a breed or lines known to be at risk for DCM or if your dog is showing symptoms. In those cases, please consult a qualified veterinary professional.

References

[1] This reference describes what DCM is: https://www.vet.cornell.edu/hospitals/companion-animal-hospital/cardiology/canine-dilated-cardiomyopathy-dcm

[2] This reference is where I obtained the market share data for grain free dog food: http://cdn2.hubspot.net/hubfs/2168539/2016_Pet_Food_Trends_and_APM_Brands.pdf?t=1484853966701

[3] This reference is a special topic review (a comprehensive survey of existing peer-reviewed research on a topic; considered a reliable source of current understanding) about what is known to date regarding DCM and diet in dogs: Wilfredo D Mansilla, Christopher P F Marinangeli, Kari J Ekenstedt, Jennifer A Larsen, Greg Aldrich, Daniel A Columbus, Lynn Weber, Sarah K Abood, Anna K Shoveller, Special topic: The association between pulse ingredients and canine dilated cardiomyopathy: addressing the knowledge gaps before establishing causation, Journal of Animal Science, Volume 97, Issue 3, March 2019, Pages 983–997, https://doi.org/10.1093/jas/sky488

[4] This reference is a review (a comprehensive survey of existing peer-reviewed research on a topic; considered a reliable source of current understanding) about what is known to date regarding DCM and diet in dogs: Diet-associated dilated cardiomyopathy in dogs: what do we know? Lisa M. Freeman, Joshua A. Stern, Ryan Fries, Darcy B. Adin, and John E. Rush, Journal of the American Veterinary Medical Association 2018 253:11, 1390-1394 https://avmajournals.avma.org/doi/full/10.2460/javma.253.11.1390