There is a thing your dog used to do coming in from the rain. The full-body shake, starting at the ears, traveling all the way down, ending with the tail. He did it after baths too, and after long naps on the hard floor.
You can’t say when it stopped. That is ordinary. Nobody notices the moment a thing stops happening. We notice things that start.
Being told that dogs hide pain tends to land as an instruction: look harder. This is about why that instruction fails, and what turns out to be within reach instead.
Quick answer
Part of it is that dogs mask pain, which veterinary medicine explains as evolutionary and treats as consensus rather than as settled fact. The bigger part is that long-term pain has no single sign that gives it away, and the changes it does produce fit a dozen other explanations. Owners misjudge how he is doing, and so do specialists: in one trial, board-certified surgeons judging by eye read roughly four dogs in every ten as better than an instrument did. What does reach you is the impact of the pain rather than the pain itself, usually as a task your dog still completes by a different route than he used to. Write those changes down as changes, take them to your vet, and let the exam do the interpreting.
Why do dogs hide pain?
The explanation that veterinary pain medicine reaches for is evolutionary. Tamara Grubb, a veterinary anesthesiologist who teaches pain assessment, puts the barrier plainly: because of an evolutionary need for survival, animals hide pain from humans and from other animals (Grubb, 2023). That is a clinical consensus about why the job is hard. It is not an experiment with a result, and the difference is worth keeping in view.
The bigger reason is what long-term pain looks like from the outside. Acute pain, from surgery or an injury, is more likely to be recognizable.
Long-term pain is different, and the criteria the Merck Veterinary Manual lists for assessing it read like a list of nothing in particular: low activity, less grooming, a smaller appetite, weight loss. Those are not specific signs of pain. They point to a problem that needs a diagnosis, and stop there (Allweiler, Merck Veterinary Manual). Owners do notice things, and no single one of them proves much on its own.
So the difficulty is structural. Chronic pain has no signature move, which is a large part of why it gets read as age instead. That reading has a guide of its own in My Dog Is Slowing Down. Is He in Pain?.
How good is anyone at seeing it?
The whole system runs on a person’s read of an animal. Grubb makes the point about acute pain scoring, where it is easiest to see: no scoring system is perfect, because we rely on a human’s perception of what the animal is feeling, or of what the animal is trying to hide (Grubb, 2023). For chronic pain she puts the owner in that same position, as the instrument.
We have one measurement of how wide the gap runs. In 2012, Conzemius and Evans looked at 58 dogs with lameness from osteoarthritis, all of them in the placebo arm of a blinded trial of an anti-inflammatory drug, comparing what people reported against force platform gait analysis, which registers how much weight a dog actually puts on a leg.
Owners overstated improvement for 56.9% of the dogs, either calling one better when the plate found him unchanged or worse, or calling him unchanged when the plate showed him worse. The specialists came out somewhat better, though the study never tested that difference directly: the board-certified surgeons examining the same dogs read them as better than the plate did for 39.7% of them at a walk, 44.8% at a trot, and 43.1% when feeling the joint for pain. Both groups drifted further from the plate as the trial went on, scoring the dogs as more improved on day 42 than they had on day 14 or 28 (Conzemius & Evans, 2012).
The useful half of that: this is what happens whenever a human eye is the instrument. It also gives you a fair picture of the odds your vet works against in a short appointment, with a dog who finds the room stressful. That is the argument for handing over something more durable than an impression.
From the clinic side, Grubb names the same reflex, that owners often mistake signs of pain for “just getting old” (Grubb, 2023). That explanation sits closest to hand, it fits everything you have seen, and it asks nothing further of anyone.
The limit: the study measured judgments of change during a drug trial. It does not tell us how often an owner misses pain in the first place, which is a different question and a harder one to put a number on.
What Grubb says about you cuts the other way, and both things hold at once. Owners know the normal behavior of their pets better than anyone, and they see them somewhere other than the stressful environment of a hospital (Grubb, 2023). You have the only long view of this dog that exists, and it comes with a margin of error nobody has managed to close.
What shows up instead?
If pain itself is hard to see, and the eyes doing the seeing are unreliable, something has to be left.
Grubb’s answer reframes the job. Owners generally have difficulty recognizing pain itself, but they can recognize the impact of pain (Grubb, 2023). You catch the dent, rather than the thing that made it.
That dent usually looks like a task still getting done, by a different route. A dog who used to fold into a down in a single movement starts doing it in two, front end first. He still sleeps on the bed, except that the ottoman became a step somewhere along the way. Turning around in the hallway used to happen on the spot, and now it takes a wider arc.
Some of that will be nothing. Dogs change how they do things for plenty of reasons, and one new habit on its own means very little. What is worth recording is the shape of it: the same task, a new route.
Write those down as changes, not as conclusions. “He lies down in two stages now, started around June” is worth bringing to an appointment. “He’s in pain” is a diagnosis, and it is not yours or mine to make. What the exam does with your observation is a separate job.
Some of what changes shows up in temper rather than in movement, and some arthritis never produces a limp at all. Neither of those is a verdict you reach at home either.
What about giving him something in the meantime?
Once you start noticing the changes, somebody will offer you a shortcut. Often it is well meant, from a relative or from a forum thread at eleven at night: a baby aspirin is fine, just a small dose, he’s a big dog.
It is not fine. Do not do it.
Naproxen is the clearest illustration of why. Its half-life in a dog runs about 74 hours, which the manual traces to the way the drug keeps cycling between his liver and his gut instead of clearing out (Hovda, Merck Veterinary Manual). Three days after the tablet, half of it is still in him.
Aspirin, ibuprofen and acetaminophen each carry their own version of the problem, and the margins are not the same from one to the next. None of them has a household dose worth guessing at, and about ibuprofen the manual is blunt: it is no longer recommended in dogs. What is on record across the four runs to stomach ulcers, kidney failure and liver injury, and for each of them the route there can be a small dose repeated rather than a large one taken once (Hovda, Merck Veterinary Manual).
Your vet can prescribe something that works for him, which is the shorter road anyway.
What do I do with what I noticed?
Book the appointment, and bring the changes with you in the words above: what he used to do, how he does it now, roughly when it shifted. A couple of weeks of that is plenty.
Written beats remembered, for the reason the study gives. In that trial, the distance between what people reported and what the plate measured grew wider the longer it ran, and the people reporting were paying close attention. A longer version of how to keep that note, and what else belongs in it, sits in My Dog Is Slowing Down. Is He in Pain?.
One kind of change does not belong in the note at all.
The test the eye can’t pass
The shake may come back tomorrow, or it may not. It was never going to settle the question for you.
Watching harder was not the fix. The thing you were watching for makes no appearance, and where there is something to see, the eye reading it drifts anyway, in a clinic as much as at home. You can stop grading yourself on it.
What changes tonight is smaller than an answer, and more useful. You start keeping a record where you were keeping an impression. The shake goes in it, with your best guess at when it went, and so does the next thing you catch him doing by a new route.
If something has been hurting, he has been managing it quietly, because that is what dogs do. Putting him in front of someone who can examine him is the part that was always yours.
Common questions
Is my dog hiding it from me on purpose?
No. It is older than he is, and it is not aimed at you. Veterinary pain medicine explains the habit by an evolutionary need for survival shared across animals, which is where the field has landed on the question rather than something an experiment settled.
How far back should I try to remember?
Not far, and do not strain for it. A guess is fine: “sometime in the spring” is a useful thing for a vet to hear, and it is more honest than a date you reconstructed. What earns its keep is what you write down starting today, because that part does not have to survive being remembered.
So should I stop trusting what I see?
Trust what you see. Hold what you conclude a little more loosely, and for a little less time. Nobody else watches him the way you do, which is why your vet wants your account of him. What the numbers argue against is letting that account sit unexamined while you narrate it to yourself.
Sources
- Allweiler S. Recognition and Assessment of Pain in Animals. Merck Veterinary Manual, rev. 2023/2024. Link
- Conzemius MG, Evans RB. Caregiver placebo effect for dogs with lameness from osteoarthritis. J Am Vet Med Assoc. 2012;241(10):1314-1319. Link Figures here follow the paper’s Results section. The published abstract transposes them.
- Grubb T. What Are They Hiding? Pain Assessment in Dogs and Cats. Continuing education, 2023. Link
- Hovda T. Toxicoses From Human Analgesics in Animals. Merck Veterinary Manual, rev. 2024/2025. Link