Oil, Treat, or Pellet? A New Dog Study Says the Format Changes What Shows Up in the Blood

The same hemp, three different ways — and the dogs' blood didn't look the same

If you've ever stood in front of a shelf trying to decide between a hemp oil, a soft chew, and a food topper, you've asked a fair question that the industry mostly answers with marketing: does it actually matter which form I pick?

A study published this spring in Frontiers in Veterinary Science took a real swing at that question. A research team at Kasetsart University in Thailand gave 18 healthy dogs cannabidiol in three completely different formats for 30 days, then looked at what happened in their blood at the protein level. The formats did not produce the same picture.

That's genuinely interesting. It's also a long way from telling you which product to buy — and the honest version of this story is the one worth reading.

What the researchers actually did

The team enrolled 18 healthy mixed-breed dogs, aged 2 to 5 years, with a mean body weight of 18.53 ± 2.95 kg. After an acclimatization period of at least 25 days, the dogs were split into three groups and each group received cannabidiol daily for 30 days in a different form:

• Hemp by-product feed pellets — 18 mg of CBD blended into a 300 g daily ration
• CBD-infused oil — 24 mg per dog per day
• A semi-solid treat — 28 mg per treat, one per day

Worth noting how those numbers came about: the researchers set out to give every group the same 24 mg daily dose. When they verified the actual cannabidiol content by HPLC, the doses came back different — the pellets delivered less than intended, the treat delivered more. That mismatch wasn't the design; it was a finding in itself.

At day 0 and day 30, the researchers ran label-free quantitative proteomics on the dogs' serum. In plain language: instead of measuring one or two markers, they used mass spectrometry to take a wide survey of which proteins were circulating and in what amounts, then compared the "before" and "after" pictures for each group. They quantified 86 proteins in the treat group, 64 in the oil group, and 63 in the pellet group.

The paper reports that all dogs maintained a normal health status throughout the study. The work was funded by Kasetsart Veterinary Development Funds and the university's research institute — not by a supplement company — and the authors declared no commercial conflicts of interest. That's worth noting in a field where a lot of the published research is industry-funded.

What they found

The headline is the one in the title: the response was formulation-specific. When the researchers ran a multivariate analysis, the feed-pellet group clustered separately from the oil and treat groups — the three formats did not simply produce more or less of the same effect, they produced different patterns.

One change did show up in every group: upregulation of the hemoglobin subunits HBA and HBB. Six KEGG biological pathways were also shared across all three.

The lipid-transport proteins are a more interesting case, and worth being precise about. The authors summarize their results as a shared core response that includes downregulated apolipoproteins — but their own results section is more granular than that summary. The pellet group showed APOA4 and APOC3 decreasing. The oil group showed APOC1, beta-2-glycoprotein I, and APOH decreasing. The treat group showed no apolipoprotein decreases at all — the proteins that fell there were hepatocyte growth factor activator and metalloproteinase inhibitor 1. So a lipid-metabolism signal does appear across the study, but describing it as identical in all three groups goes further than the data does.

We're pointing that out because we'd want someone to point it out if we did it. Reading a study means reading past the summary paragraph, including when the summary belongs to researchers we respect.

Beyond that shared core, the two higher-exposure formats diverged. In the oil and treat groups, the researchers reported increases in proteins associated with vascular integrity and tissue scaffolding. The treat group alone showed changes in markers the authors describe as immunomodulatory and cytoprotective — including clusterin — and uniquely activated 22 additional pathways the other two groups did not.

The authors' own interpretation of why the pellet group looked different is a hypothesis, not a finding: they suspect reduced bioavailability, meaning less of the cannabidiol may have survived the feed-processing and digestion route to reach the bloodstream. They are explicit that they could not confirm this, because they did not measure CBD levels in the blood or quantify the cannabinoid content before and after the pellets were processed.

What this study does not show — and it matters

This is where most coverage of a study like this goes wrong, so let's be direct about it.

These were healthy dogs, and nothing clinical was measured. Nobody scored comfort, mobility, behavior, or quality of life. A shift in a circulating protein is a biological signal, not a benefit. Some of these proteins move for reasons that have nothing to do with wellness, and the paper does not claim otherwise.

There was no untreated control group. Without a group of dogs that got nothing, you cannot fully separate the effect of the cannabinoid from the effect of 30 days passing, a change in diet, or the season.

The diets themselves weren't identical. To keep calories and protein matched across the three groups, the rations had to be formulated differently — and the authors state plainly that these compositional differences "cannot be excluded as potential contributors to the observed proteomic effects." Some of what separates the pellet group may be the feed rather than the cannabidiol in it.

The samples were pooled. Combining blood samples within each group is efficient, but it erases individual variation — and individual variation in how dogs metabolize cannabinoids is one of the most consistent findings in this whole field. Your dog may not be the average dog.

Eighteen dogs is small, and 30 days is short. The authors say so themselves, and describe their results as "foundational molecular frameworks that require further validation in larger, unpooled clinical cohorts." That is a research team being appropriately careful about its own work, and it deserves to be repeated rather than sanded off.

And because the doses weren't matched — 18, 24, and 28 mg — format and amount are partly tangled together. Some of the difference between groups may simply be a difference in how much they got.

Where this sits next to the rest of the evidence — including ours

We grade every study we write about on the same eight-level ladder, and we apply it to our own research too. That's the whole point of having a ladder.

This Kasetsart study is Level 3. HempMy Pet's two university studies also sit at Level 3 — a 90-day study in 30 dogs with chronic joint discomfort, conducted with Dr. Robin Downing, DVM, alongside Lori Kogan, Ph.D. and Peter W. Hellyer, DVM of Colorado State University, and an equine study at Tarleton State University with Dr. Kimberly Guay, Ph.D. that reported reduced measures of stress during transport.

Level 3 is not the top of the ladder, and we're not going to pretend it is. But it's a meaningful distance up from where most of this industry actually stands: Level 8, anecdotal reports and testimonials. When a brand tells you something works and the evidence behind it is a review section, that's Level 8. We think you should know the difference, and we think you should hold us to it too.

So what does a pet owner do with this?

Three practical takeaways, none of which require you to change anything today.

Format is a real variable, not just a convenience choice. The most useful thing this study contributes is evidence that how a cannabinoid is delivered changes what the body sees. Owners have long been told format is purely about preference — "whatever your dog will take." This suggests it's worth a little more thought than that. We've written a fuller practical comparison of CBD oil versus treats for dogs if you're actively choosing between them.

Heavily processed delivery routes deserve scrutiny. The format that looked most different here was the one blended into a processed feed pellet, and the authors' leading explanation is reduced bioavailability. That's a hypothesis, not a conclusion — but it's a reasonable prompt to ask any company how their product is made and what survives the process. The carrier the extract sits in is part of that question too, which is why we've written separately about why we use virgin coconut oil as a carrier.

Milligrams on the label are only half the story. This study is a clean illustration: the researchers aimed for 24 mg across the board and, once they actually measured it, got 18, 24, and 28. If the delivery format influences both what's in the product and what reaches the bloodstream, then comparing two products by the number on the front of the bottle is comparing incomplete information. Ask for the batch-specific certificate of analysis instead — ours are all published on our lab results page, and any company should be able to hand you the same thing.

And the honest bottom line: this study is a signal that the question is worth asking, not an answer to it. We'll cover the follow-up work when it lands — including if it complicates the picture.

Frequently asked questions

Does this study mean CBD oil is better than treats for my dog?
No. The study did not compare how well the formats worked, because it did not measure any outcome you could observe in a dog. It measured protein patterns in the blood and found they differed by format. Which of those patterns is preferable — or whether that question even makes sense — is not something an 18-dog, 30-day study in healthy animals can answer.

What is a "serum proteome," in plain English?
It's the full set of proteins circulating in the blood at a given moment. Rather than testing for one specific marker, proteomics surveys thousands at once and reports which ones went up or down. It's a wide-angle look, useful for spotting where to investigate next — and it's exploratory by nature, which is exactly why the authors framed their results as a starting point.

Is it safe to give my dog hemp products every day for a month?
That's a conversation for your veterinarian, who knows your dog's history and any medications they're on. What this particular study reported is that the dogs maintained normal health status over its 30 days. That is one small study in healthy young adult dogs, not a general safety guarantee, and safety questions in this category are still actively being researched — the FDA's Center for Veterinary Medicine is a reasonable place to follow the regulatory side.

Why do you grade studies instead of just saying the science supports your products?
Because "studies show" is doing a lot of unearned work across this industry, and pet owners deserve to know whether the study in question was 18 healthy dogs for 30 days or a systematic review of thousands. Grading forces us to be specific — including when the honest grade is lower than we'd like.

Where can I read the original paper?
It's open access: Formulation-specific dose–response in the serum proteome of healthy dogs following cannabidiol administration, published in Frontiers in Veterinary Science by researchers at the Kasetsart University Faculty of Veterinary Medicine. For the record, Frontiers published a correction in June 2026 reassigning which author was listed as corresponding author. No data, results, or conclusions were changed.

HempMy Pet products are hemp-derived supplements. They are not intended to diagnose, treat, cure, or prevent any disease. Always talk with your veterinarian before adding any supplement to your pet's routine, particularly if your pet is taking medication.