Introduction
Given the media frenzy the article has kicked off, it is unlikely very many people are unaware of a recent study reported in the journal PLOS One that claims to show vegan diets are healthier for dogs and cats than diets containing meat.
Knight A, Huang E, Rai N, Brown H. Vegan versus meat-based dog food: Guardian-reported indicators of health. PLoS One. 2022 Apr 13;17(4):e0265662.
Dramatic headlines have appeared in many major news sources proclaiming that vegan diets are “linked to better health” or “are healthier and safer” than conventional diets. Getting beyond the headlines, one may even read that this study demonstrates raw diets to be healthier than conventional foods, though that is less emphasized in both the original article and the media coverage.
I’ve written about vegan diets for dogs before, and indeed this article cites my column on the subject, though only to dismiss my claim that vegan diets “should not be recommended” as “without evidence.” This is not entirely accurate, since my article did cite the limited evidence available, and I did not actually recommend against vegan diets for dogs, though I did conclude they were a bad choice for cats. My actual claims were:
There is no evidence vegetarian diets have health benefits for dogs and cats, and no real reason to believe they should, based on the physiology and nutritional requirements of these species.
Dogs are omnivores that are able to eat both plant and animal foods, and in theory, they should be able to thrive on vegetarian or vegan diets. However, these diets must be carefully formulated, and many commercial vegetarian dog foods do not appear to be nutritionally adequate. There is also little reliable research evidence showing dogs can remain healthy when fed only a vegan diet. Given the unexpected health problems seen with theoretically adequate grain-free diets, we should be cautious about the potential risks of vegetarian formulations for dogs until there is better and more evidence showing their long-term health effects.
Cats are clearly obligate carnivores with nutritional requirements that are unlikely to be effectively met by vegan diets. Such diets offer only risks and no benefits for cats and should be avoided.
In any case, I am sympathetic to many of the potential environmental and health benefits of plant-based diets, and I have been a vegetarian, though not a vegan, for almost 20 years (technically, an ovo/lacto/pescatarian, though realistically I don’t like fish and so eat very little of it). However, the evidence for benefits of plant-based diets for dogs and cats is far weaker than that for humans, which is itself often nuanced and not always conclusive. And I believe in following the evidence wherever it leads, even when it doesn’t support my personal beliefs or practices.
So, does this paper change the game in terms of showing real health benefits to raw and/or vegan diets? Spoiler, but not really! Let’s look at the actual findings a bit closer.
The Study- Results
The study reported the following statistically significant differences between “conventional,” raw, and vegan diets (leaving aside the fuzzy definitions of these, the ubiquitous feeding of unidentified treats to all pets, and the lack of clarity about how strictly feeding practices corresponded to each category).
- Dogs fed raw and vegan diets were less likely to have had multiple veterinary visits in the year observed.
- Dogs fed raw and vegan diets were less likely to have been given medications
- Dogs fed raw and vegan diets were less likely to have been transitioned onto a therapeutic diet
- Dogs fed raw and vegan diets were less likely to have an owner-reported veterinary assessment of poor health status
- Dogs fed raw and vegan diets were less likely to have an owner reported assessment of poor health
- Dogs fed raw and vegan diets were less likely to be reported to be “unwell,” and when unwell dogs on raw diets had fewer reported disorders than dogs on conventional diets.
- There were differences in the occurrence of specific health conditions between diet groups as illustrated in this figure and table:
- Dogs fed raw and vegan diets were less likely to have had multiple veterinary visits in the year observed.
- Dogs fed raw and vegan diets were less likely to have been given medications
- Dogs fed raw and vegan diets were less likely to have been transitioned onto a therapeutic diet
- Dogs fed raw and vegan diets were less likely to have an owner-reported veterinary assessment of poor health status
- Dogs fed raw and vegan diets were less likely to have an owner reported assessment of poor health
- Dogs fed raw and vegan diets were less likely to be reported to be “unwell,” and when unwell dogs on raw diets had fewer reported disorders than dogs on conventional diets.
- There were differences in the occurrence of specific health conditions between diet groups as illustrated in this figure and table:
All-in-all, these results would seem to be pretty bad news for conventional diets (whatever that means) and pretty good news for feeders of raw and vegan diets. Anyone reading this, however, can probably sense the “but” coming, so let’s get into the caveats.
The Study-Limitations
The biggest limitation (and boy is it a big one!) is that all the data were collected from online surveys of pet owners. Everything from the diet fed to the health status and even the reported veterinary assessment of health all relied entirely on the responses of pet owners who found the survey online and took the trouble to complete it. None of the facts, not the feeding practices nor the health of the pets, were verified objectively in any way. Right from the start, then, that makes this a study of what some small subset of pet owners believe about their pets’ diet and health, not about the actual feeding and health of these animals.
It is not difficult to find reasons to be concerned that these responses may not reflect reality, or even the opinions of other groups of pet owners. The respondents were overwhelmingly female (92%) and European (86%), which is a pretty narrow population to survey. They were, of course, also in a socioeconomic class inclined to participate in online pet health surveys, which is also not representative of many other pet owners.
More concerning, 22% of the population of owners were themselves vegan. This is a group likely to have strong beliefs and biases about plant-based vs meat-based nutrition and health, so it would be, frankly, shocking if they did not report that plant-based diets were healthier than other diets. 13% of these owners fed their pets vegan diets and 33% fed raw meat, indicating nearly half fed unconventional diets not usually recommended by veterinarians or veterinary nutrition specialists. Such a group is clearly a population biased in favor of the kind of outcomes reported in this study, and given the study only shows us the opinions of owners, not actual objective data about health and diet, the outcomes are simply a fancy way of reporting what people in this group think.
Other studies in other populations have reported lower rates of veganism (5.8%) and of feeding vegan diets to pets (1-3%), which emphasizes that this study surveys a very specific, narrow group of owners. A review of previous owner surveys also shows more nuanced and variable feeding practices reported in these studies, again suggesting this paper may not be applicable outside of the specific population included in the survey.
In terms of potential sources of bias beyond the survey population, it is worth pointing out that the lead author is a consistent advocate for plant-based diets, for humans and pets, on environmental and ethical grounds. The study was also funded by a plant-based diet advocacy group. As I have discussed in detail in previous articles on conflict of interest, these facts do not indicate the research is fraudulent or inaccurate, nor are they a justification for ignoring the claims, arguments, and evidence provided in this paper. They are, however, a reason to consider carefully the potential for uncontrolled unconscious bias in the design, conduct, analysis, and reporting of the study. Given that the study itself was essentially a survey of subjective opinions, methods to control such bias are minimal, so the results have to be viewed in that context.
The authors, to their credit, do acknowledge some of these limitations. They call out the fact that, for example, fewer reported veterinary visits in dogs fed raw diets may be due to the fact that feeders of raw diets are often skeptical of conventional veterinary medicine and less likely to seek care rather than to any actual difference in health status. The same logic, of course, applies to the frequency of vet visits for dogs fed vegan diets, and to the reported use of medicine and therapeutic diets by owners who have a clear preference for unconventional health practices to begin with.
Conclusions
So what does this study mean? Overall, it means that the particular population of pet owners surveyed believes that feeding raw and plant-based diets are associated with better health in their pets. They also believe that their veterinarians think their pets are healthier (though whether these vets actually believe this is unknown). And these owners report less use of veterinary medical services, though whether this means their pets are healthier or simply that they try harder to avoid taking their pets to the vet is also unknown.
Like previous studies relying on owner surveys and both conducted and funded by folks with strong a priori opinions about diet and health, this is a useful insight into such beliefs. It is not compelling or probative evidence for actual health effects of different feeding strategies. Sadly, the media coverage of the paper rarely recognizes or emphasizes this.
As the authors themselves suggest, though with little evident enthusiasm, controlled studies with objective measures of outcome and more defined and verified feeding practices are required to draw any meaningful, actionable conclusions about the healthiest feeding strategy for our pets. I am neither for nor against vegan diets for dogs, and I am even open to reversing my objection to feeding vegan to cats or raw diets to cats or dogs if strong evidence is generated that these are safe or beneficial practices. However, regardless of the difficulties in funding and conducting the necessary research, we are not justified in making confident claims about the health impact of raw, vegan, or conventional diets without it.
It’s interesting that some off the questions had to do with frequency of vet visits and use of medications and therapeutic diets. It’s entirely possible that people who fed raw or vegan diets had less trust in veterinarians and that that explained fewer visits and less use of conventional veterinary medicine. Lots of confounders in this study for sure.
Thanks for the review. I heard this story on BBC and was rather skeptical of the claims made. A study based on no objective evidence is not very convincing.
2 problems i have with this discussion:
Some studies have found nutritional deficiencies in commercial vegan diets therefore even properly formulated products are having issues (I’m guessing nutritional interactions such as plant based anti-nutrents are not being in-depthly researched).
The second is the premise that dogs are omnivores to begin with as most ecological sources discuss them as mainly preditory scavengers (borderline hyper carnivores). As a subspecies of grey wolf that can interbreed with grey wolves with completely fertile young, at what percentage wolf is a wolf-dog suddenly not an omnivore anymore? There are a couple of gene’s that help them digest certain carbs that’s true, but how does this suddenly mean they are capable of having only plant based foods? (Most candids can’t do this, even the most vegetarian dog like species (maned wolf) eats 50% vegetation 50% animal based foods) And if the only difference between cats and dogs is this one type of gene and the need for taurine then why is vegan cat food suddenly dismissed when they will have the same nutritional considerations within this type of diet but with additional taurine?
I’d very much like to know what was in the vegan diets fed to dogs, and particularly to cats – I’m not expecting an answer here obviously. Especially in the case of cats it is difficult to see how a vegan diet could be healthy and at all ‘natural’. I can’t help wondering how much the cats supplement what they are fed by their owners with hunted food.
The current consensus is that dogs have co-evolved with humans, and this has involved significant adaptations to a more omnivorous diet. The adaptations are a bit more extensive than a couple of genes, and dogs have been thriving on essentially human diets for a long time. Whether or not there are health benefits or risks to properly formulated vegan diets isn’t truly known, and can’t be without proper feeding trials, but the imperfect evidence we have makes it at least plausible that they should b able to do well on such diets.
One thing I’ve noticed in discussions of vegan cat food is that vegans will often point to the nutritional label, saying “Look, it has everything cats need” without acknowledgement that even if it’s printed on the paper it might not be as bioavailable to cats as the nutrients in meat sources. As you noted, there is a strong ideological bias coloring their understanding of pet food, their observations, and their reporting. And I say this as a completely committed vegan myself. Veganism is an expression of my values, but that’s the thing: they’re MY values – I’m not going to impose those values on my cats unless our vets assure us there is a vegan alternative with more than just anecdotes and shaky owner reporting surveys to back it up.
“ubiquitous feeding of unidentified treats to all POETS” (emphasis mine). I do have a book of poems by dogs, so maybe identification of dogs as poets is justified. 🙂 🙂
I’m all for seeing poets well fed! 😉 (typo fixed)
Um… why do you think dogs are omnivores? The evidence clearly states that dogs are in fact carnivores. Why?
They lack amylase in their saliva, which breaks down carbs. ALL omnivores have this. Dogs don’t.
Dog’s jaws don’t move side to side like an omnivore’s. They only move up and down, and they are not meant for grinding plants.
A dog’s liver produces it’s own vitamin C. The only deficiencies an omnivore would get would be a vitamin c deficiency if it ate only meat. Dogs wouldn’t get that deficiency.
If you need more evidence, please feel free to ask.
Oh, and I’d love to hear your opinions on why you believe dogs are omnivourous??
After 10-15,000 years of eating human leftovers, and a couple of centuries of intensive breeding, dogs have significant anatomic and physiologic differences from wolves. These include adaptations in dentition, GI function, and expression of genes related to starch digestion and glucose absorption. They are phylogenetically carnivores but functionally omnivores, and millions of them live long, happy lives on omnivorous diets. This is a classification widely accepted by veterinary nutritionists and other scientists with relevant expertise.
Now classifications like this are not always bright, hard lines. Obligate carnivores may not need or be able to digest plant material, but there is a wide range of capacities for eating both animal and plant foods. Humans can manage on a plant-only diet more easily than dogs, and giant pandas (also carnivores, phylogenetically) can do so better than humans. The idea that dogs descend from wolves and so must be obligate carnivores is simplistic and inaccurate. You might want to look at several previous posts on this topic- 1 2, 3.
I see your point. However, there is evidence that shows dogs have not “changed” since the beginning of their domestication (by the way, dogs have only been domesticated for about 5,000 years or so). I do not even believe dogs descended from wolves.
You see, dogs still have these carnivorous features today (such as, but not limited to, a lack of amylase, a short GI tract, a specific tooth structure, no amylase in the saliva) are there.
I don’t believe that dogs are carnivores because they “decended from wolves”, nor do I believe that they are carnivores simply because they fall under the order carnivora. I believe that they are carnivores because all their physical features state so.
I don’t, however, believe that dogs are obligate carnivores (having to eat meat to survive), but instead are faculative carnivores (survives primarily on animal tissue). Did you catch the word primarily. We can feed them veggies and grains for as long as we want to, but it’s not going to change their features.
Oh, and while we’re on the subject of dogs and cats, why do you think cats are “still” obligate carnivores, after eating what humans do, and dogs aren’t?
I’d like to hear your opinions on this
Oh, and why do you think most dogs have long, happy lives? According to genetics, dogs should be living about twenty years, when the average lifespan is about 10-12 years now. Obesity, kidney failure, and cancer are at an all time high in pets. Is it coincidence, or is it the factors of eating more plants than they ought to?
You see, carbohydrates, when consumed in excess amounts, can cause obesity; in carnivores, plant protein has been linked to kidney failure. I’m not so sure if cancer weighs into the plant scale or not, I think that is more likely due to the amount of chemicals in the water in some areas, and the amount of artificial substances in the food (I wouldn’t be surprised if it has some connetcion, though, to not eating enough meat).
Well, that’s just something to think about.
I do not even believe dogs descended from wolves.>>>
what animal do you think dogs descended from? wolves will breed with coyotes and dogs with wolves. Didn’t they even find out what kind of wolf dogs came from the dna?
So many misconceptions….
“According to genetics, dogs should be living about twenty years” Any evidence for this? The biggest predictor of lifespan is body weight, and toy breed dogs often live close to 20, but there is not way of using some vague “genetics” to predict lifespan, so this is not a reasonable claim to begin with.
“Obesity, kidney failure, and cancer are at an all time high in pets.” Obesity is, and it is a consequence of overfeeding, not carbohydrates per se.
Any evidence for the claim that the incidence of kidney failure is increasing? Any evidence this has anything to do with diet (as opposed to dogs living to be older because they are less likely to die young from infectious disease and trauma, which is the more likely explanation for changes in the incidence of age-associated diseases)? Same questions for neoplasia.
“carbohydrates, when consumed in excess amounts, can cause obesity” So can fat and much more easily than carbohydrates since it is more calorie dense. So can protein. Excess calories, no specific macronutrients, cause obesity.
‘in carnivores, plant protein has been linked to kidney failure” Evidence? The best research on diet and kidney disease actually shows that moderate protein commercial diets prolong life expectancy and delay progression, and this has not been shown for low-carb diets.
“likely due to the amount of chemicals in the water in some areas, and the amount of artificial substances in the food (I wouldn’t be surprised if it has some connetcion, though, to not eating enough meat).”
All of this is just vague hand-waving showing you don’t actually know anything about nutritional risk factors for cancer, you just have a knee-jerk prejudice against things you think ar e’unnatural” such as “chemicals” (which, of course, means all food since everything is made of “chemicals”) and ‘artificial substances,’ (here’s an article detailing the evidence concerning dozens of these, which debunks the idea that they are inherently harmful).
“(by the way, dogs have only been domesticated for about 5,000 years or so)”
Actually, much longer. Genetic evidence is more consistent with about 15,000 years.
‘I do not even believe dogs descended from wolves.’
Well, they absolutely did, so your ‘belief” contradicts the evidence and opinions of virtually every expert in the subject.
“all their physical features state so.” Which ignores all of the functional and anatomic changes brought about by domestication, including increasing ability to digest starches and the obvious anatomic and behavioral changes in many modern breeds. Have you looked at the dentition of a chihuahua or a pug lately? Not suitable for a predatory lifestyle at all!
“faculative carnivores” Yes, but you misunderstand what this means. “Facultative carnivores are those that also eat non-animal food in addition to animal food. Note that there is no clear line that differentiates facultative carnivores from omnivores; dogs would be considered facultative carnivores.”
“We can feed them veggies and grains for as long as we want to, but it’s not going to change their features.” Untrue. Natural and artificial selection have very much changed dogs, and one element in this is that those who were better adapted to eating the food wee provided them with survived and reproduced more successfully than those less well-adapted to it, which is how evolution works. human feeding and selective breeding are selection pressures that have made dogs dogs instead of wolves, and that has influenced their dietary needs as much as any other aspect of their being.
‘why do you think cats are “still” obligate carnivores, after eating what humans do, and dogs aren’t?” For nearly all the time they have associated with us, cats have not eaten human food primarily but have hunted the rodents that ate our food. They have been much less affected by selection, natural and artificial, in their association with us because the main thing we wanted from them was to hunt and kill things for us, which they were already doing. This is quite different from dogs, who were selected for guarding, herding, hunting, copmanionship, and many other roles that did not involve predation.
Having fed large packs ( up to 17) of mainly small dogs of my own for the past 35 years and also owning a boarding kennel, my own approach to feeding is variety is the spice of life. Nearly all of my dogs have exceeded the average life expectancy of their breeds and remained well until old age finally caught up. Most have never had a vet visit for illness, and only visited vet for their puppy vaccinations,.
The base is high grade raw and very varying meat and offals but also various good quality dried foods, some vegetarian, some not, which I usually mix in. Occasionally some herb mixes ( yes I know ??) sardines, kefir ( goat or cow) a little rice, barley or oats depending on weight etc. I have found that some dogs need the addition of carbs to keep weight on and some need virtually none expect for that found in small amounts of veg. and remain the perfect weight.
Like humans, variety is the spice of life and their nutritional needs will be met. My very oldies receive high dose Vit B complex which really does boost their activity levels. Yes Skeptvet it’s anecdotal. Some old dogs get the odd shaky episodes and this definitely reduces or eliminates them.
Don’t get hung up on any particular diet, use high quality foods whether raw, wet or dry and give them variety. Not all very old dogs can stomach raw and need their food lightly cooked but most seem to do well on it until the end.
I did know one dog that refused to eat meat and lived very well on vegetarian diet
successfully rearing several litters.
My Veterwn show dogs are renowned for their condition and longevity and so I can only conclude that how I feed is working.
Skeptvet, as a long time appreciative reader of yours, I value your writings and very rarely, could make a reasonable argument against any of them, nice work!
I grew up in suburbia, in an era before the advent of pet food superstores or even plain jane Walmart stores, except in more rural areas. The choices of dog food was very limited, usually buying from supermarket, or occasionally making longer trip to a rural feed store – even there, the choices were essentially limited to the large manufacturers. Later on, Eukanuba and Iam’s, if you could get them, were considered the most premium foods.
Very few friends and aquaintances owned dogs, especially those kept in the house. We always considered dogs to be basically carnivores, and I’m pretty sure, most scientists back then also. We knew the kibble was largely corn based, but accepted it as a must use, as money usually tight. As an occaisonal, we would treat to a canned food, which I believe was almost exclusively horsemeat, Strongheart, Kenl-ration, etc., and somewhere along the line, Gaines Burgers, (are those still around?) I remember being so amused and astonished when certain dog kibble packaging and advertisements first began showing vegetables on them, I think it was beneful, ha.
I’m certainly not here to argue that dogs are highly carnivorous wolves, but to this day, I still can’t think of dogs as actual omnivores, at least not without caveats. The best I can do is call them “quasi” omnivores, i.e. “almost” omnivores.
My comments below are based not on your article, but on some of the comments and links you provided to chattycatty.
You quote from the book, “Facultative carnivores are those that also eat non-animal food in addition to animal food. Note that there is no clear line that differentiates facultative carnivores from omnivores; dogs would be considered facultative carnivores.” Clearly, the author states, that dogs are facultative carnivores, he doesn’t merely say that dogs are omnivores. Perhaps facultative carnivores could be a more specific kind of omnivore? I wouldn’t consider a mouse, a fox, or a human to be a facultative carnivore. Maybe the general term “omnivore” encompasses the more specific, but still fuzzy term “facultative” carnivore, but not really vice versa?
The other article/research paper you linked to; “Dietary nutrient profiles of wild wolves: insights for optimal dog nutrition?” , is a very well written thoughtful, composition., but I’m not sure you aren’t missing the point of it, or ignoring it, or? The authors seem to be questioning the more recent thinking that a dog is actually an “omnivore”. “The scientific confirmation of the absence of identical or similar specialised metabolic pathways in dogs has led many scientists to question the once-firm carnivorous classification of our domesticated dogs. This ‘omnivorous dog dogma’ has developed over the past 40 years and has found its way into authoritative scientific reference books( 12 , Reference Hand, Thatcher and Remillard13 ), nutritional concepts in pet nutrition and as a general view.” The authors argue that current scientists compare traits that are similar in cats, an obligate carnivore, and traits that are dissimilar, but most dissimilar traits are the traits that enabled a wolf to survive a feast/famine lifestyle, and also to become domesticated. (see fig. 2). Most of the other dissimilar traits (not talking about physical and behavioral), are affected by domestication and selective pressures, i.e. increased capacity for glycogen/starch digestion and for glucose uptake. I understand that the authors go on to question whether the current typical high carbohydrate foods, almost always kibble, continue to put physiological and metabolic challenges and pressures on the modern dog.
Would really appreciate your thoughts!
Thanks for you comment. As I try to always emphasize, nutrition is a complex field, and there are rarely definitive studies to answer many of the questions dog owners have because these typically are complex, time-consuming, and expensive to conduct. As a result, I often point out that I don’t think we know what the optimal diet is for dogs. We know that many individuals thrive on a wide variety of commercial and home-made diets, but whether one of these diets is clearly better than some or all of the others sand for all dogs or just some dogs is not a question we have sufficient evidence to answer with confidence.
My own completely personal suspicion is that the cooked fresh food approach will ultimately turn out to be the best balance of risks and benefits and may have net health benefits over the long term for most dogs. But this is just my best guess, and there isn’t robust data to support this.
My main concerns about the discussions that usually happen around feeding strategies for dogs is that they
1. involve a great deal more confident claims than can be justified by the evidence
2. usually involve extrapolating from theories or limited evidence from lab studies or basic physiology, which often do not represent the complexities of the whole dog in the real world
3. they are rife with simplistic dogma and rigid positions that people then simple try to justify rather than honestly interrogate with research.
With regard to your specific comments, I don’t know that the categories of omnivore and carnivore can be clearly delineated or that which category we put dogs in makes that much difference in terms of settling the question of how to feed them. They have undergone significant behavioral, anatomic, and physiologic changes form wolves due to domestication and selective breeding, and several of these changes appear to be adaptations towards a diet consisting largely of human leftovers, including much more plant material and carbohydrate than a wolf in the wild would eat. Whether this changes their category from facultative carnivore to omnivore is a question of how we define the categories, but I think it suggests that the anxiety many have about plant ingredients and carbohydrates in dog foods is misplaced. There is abundant evidence that dogs can make use of these ingredients and thrive on diets containing them. At most, we might someday see that there is some net health benefit on a population level to reducing the proportion of these ingredients in dog foods if we can do so and still make high-quality, balanced nutrition affordable and available to dog owners, and if the evidence comes to show this (as, again, I suspect it may with regard to fresh food diets), then we should certainly make such a change. But that is quite different form decrying commercial foods as “poison,” as so many do, and suggesting dramatic health benefits over short periods in indifivual dogs from change to plant-free, raw, ketogenic, or some other specific alternative feeding strategy.
Hope this answers your question?