Reading Pain in an Old or Sick Dog

Written by thecanineledger

In crisis right now? Canada and US: call or text 9-8-8.

Ask an owner of a fourteen-year-old dog how he’s doing and you will often hear a version of the same sentence: he’s slowed down a lot, but he’s fine: he’s just old.

Sometimes that is exactly right. Often it isn’t.

This is not a subtle observation that needs defending. It is written into veterinary guidelines in plain language. The World Small Animal Veterinary Association’s pain guidelines state it directly:

“However, chronic pain is often not diagnosed as it is often mistaken for ‘getting old’. Veterinarians treating animals with chronic disease should consider the potential for accompanying chronic pain, even in the absence of immediately obvious signs.” — WSAVA Guidelines for Recognition, Assessment and Treatment of Pain, 2014

If you are reading this while working out whether your dog still has a good life, this matters more than almost anything else you will read. Pain is the single category on any quality-of-life assessment most likely to have an available answer. A dog scoring badly because of untreated pain may be a dog whose pain plan needs revisiting, not a dog at the end.

1. Why this goes wrong so reliably

Three things work together, and none of them is anyone’s fault.

Chronic pain arrives too slowly to notice. WSAVA: “The behavioural changes associated with chronic pain may develop gradually and may be subtle, so that they can only be detected by someone very familiar with the animal (usually the owner).” A change spread over eighteen months is invisible from inside the household. You are comparing him to last week, not to three years ago.

“Old” is an explanation that stops the search. A 2023 study interviewed fifteen owners of dogs aged 8–17, plus eleven veterinary professionals, and collected photos or videos, with written descriptions, from sixty-one more owners. Its title is “Just old age.” Its finding: “Age-related changes were mostly perceived as ‘just old age’ by dog owners.”

One owner: “What she had was aching joints and stiffness of muscles, issues of old age.” Another: “If they are sleeping a lot because they are old, then you tend to think, ‘Well, they are alright.’” And from a veterinarian: “Some people don’t think their dog needs pain relief when, watching it walk, I’ve fairly certainly decided it probably does” (Wallis et al., Journal of Small Animal Practice 2023).

A note on that study, since we hold ourselves to describing evidence accurately: it is qualitative, UK-based, and used a small purposive sample. It establishes that this attitude exists and was common among the people interviewed. It does not tell you what proportion of all owners think this way, and we are not going to convert it into a percentage.

The conversation often doesn’t happen. From the same study: “Although VP may identify pain signs, if the owner does not mention anything, they may not point it out, due to time constraints and difficulties in convincing owners their pet may be in pain.” One veterinarian described owners treating a dog never needing the vet as a point of pride: “People tended to wear it as a badge of honour, saying ‘Oh yes, she’s 14 and she’s never been to see the vet’.”

So the owner assumes the vet would have said something, and the vet assumes the owner would have raised it.

Our note: All WSAVA quotes here are from the 2014 edition, and a 2022 edition exists.

2. Arthritis in an old dog is not the default state

Worth stating clearly, because the assumption underneath “he’s just old” is that stiffness is what happens to old dogs.

A 2024 study radiographed four joints in 1,873 dogs over eight years old. It found osteoarthritis in 39.2% of shoulders, 57.4% of elbows, 35.9% of hips and 36.4% of stifles, and the authors’ conclusion runs directly against the assumption:

“As most of the examined joints were free of OA, radiographic findings suggestive of OA should not be considered normal in senescent dogs.” — Roitner et al., The Veterinary Journal 2024

Two caveats we won’t bury. These dogs came to a university teaching hospital, mostly for orthopaedic problems. The authors name the selection bias themselves, so these are not general-population figures. And this is radiographic change, which correlates imperfectly with pain: a dog can have arthritic joints on film and not much discomfort, or the reverse.

The point stands anyway, and it is the authors’ own: arthritis in an old dog is a finding, not a fact of life.

3. What pain actually looks like

Signs below are ordered by how well evidenced they are, not by how often they appear on lists. Where something is expert opinion rather than measurement, it says so.

The strongest evidence: restlessness at night

If you take one sign from this page, take this one, because unlike almost everything else here, it has causal evidence behind it.

A double-masked, placebo-controlled crossover trial gave meloxicam or placebo to fifteen dogs with osteoarthritis (mean age 10.3 years) and measured activity objectively. On the pain medication, the dogs showed decreased night-time activity between 11pm and 6am and increased daytime activity (Gruen, Samson & Lascelles, Scientific Reports 2019). The same direction appeared in a trial of dogs with bone cancer.

Treat the pain, and the night-time restlessness improves. That is a much stronger class of evidence than a list-of-signs assertion, and it makes returning night-time restlessness one of the better signals that pain is no longer controlled.

There is also a validated, owner-completed sleep questionnaire for dogs, SNoRE 3.0, checked against polysomnography and accelerometry (Mondino et al., Scientific Reports 2023).

Honest limit: sleep disruption is well evidenced. Changes in sleeping position or chosen spot, widely listed, are not SNoRE items and we found no study measuring them. Plausible, not evidenced.

Very strong: the things he has stopped doing

These are items inside psychometrically validated instruments, which is the strongest evidence class for “this is a recognised sign of pain”:

Difficulty or reluctance rising from lying down
Difficulty or reluctance on stairs, curbs, doorsteps
Reluctance to jump: onto the sofa, into the car
Reduced walking and running
Reduced play, and reduced willingness to go out

Every one is an item in the Canine Brief Pain Inventory or the Helsinki Chronic Pain Index. WSAVA’s own list names “ease of lying, sitting, jumping up, tolerance to exercise.”

Notice these are all things a dog stops doing. There is no yelp. That is exactly why they get filed under ageing.

Strong: posture, movement and touch

From the 2022 AAHA Pain Management Guidelines:

“At rest, awkward limb positions may indicate the presence of joint pain; for example, dogs with elbow joint pain may supinate their forelimbs and flex their carpus”
“When standing and moving, dogs shift weight away from a painful limb”
“A ‘tucked up’ appearance can be an indicator of abdominal pain”
“Postural straining can be an indicator of lower urinary tract or lower gastrointestinal tract pain”

Reacting to being touched or handled where he previously didn’t is a scored category in the Glasgow pain scale.

Strong: irritability, and behaviour that looks like temperament

This one is under-appreciated and matters enormously, because it gets read as a dog “turning grumpy in old age.”

A 2020 review of behaviour caseloads found that “a conservative estimate of around a third of referred [behaviour] cases involve some form of painful condition, and in some instances, the figure may be nearly 80%” (Mills et al., Animals 2020). Signs listed include aggression, defensive behaviour, reluctance to move, house-soiling, clinginess, and compulsive licking.

Read that figure carefully: it comes from the authors’ review of 100 recent referral cases: dogs already sent to a behaviourist. It is not “a third of all dogs.” What it does say is that when a dog’s behaviour changes, pain belongs on the list of explanations.

The same paper offers a mechanism worth knowing: early warning signals “such as head turns and hard stares… have been ignored or not noticed prior to this and possibly may no longer be expressed as a result.” A dog who has learned that the quiet signals achieve nothing may skip to the loud one.

Moderate: licking

Licking or chewing at one spot is a scored category in the Glasgow scale. Separately, and less well known, copious licking of surfaces (floors, walls, carpets) is associated with gastrointestinal pain.

Weaker than you have been told: appetite

Listed by WSAVA and AAHA, but as expert assertion rather than measured association, and in an old or terminally ill dog, appetite loss is so non-specific that it tells you little on its own.

Our note: The order of these signs is ours.

4. Four things you will read elsewhere that the evidence does not support

Part of what this site promises is telling you when a widely repeated claim has nothing behind it. These four are everywhere.

“Dogs instinctively hide pain”

Usually accompanied by an evolutionary story: showing weakness marks you as prey, or gets you pushed down the pack hierarchy.

Hospice guidance says so, but the evidence for dogs is weak. The international animal hospice guidelines state: “Since animals are instinctively inclined to hide their pain…” (IAAHPC guidelines). We found no study showing that dogs hide pain by instinct. The pack-hierarchy version is dominance-theory folklore.

What is true is narrower, better evidenced, and more useful:

“Indicators of pain may be masked by behaviors characteristic of the species being observed, such as a dog wagging its tail and greeting an observer at the cage door despite being in significant pain.” — Sander et al., Frontiers in Veterinary Science 2025

Your dog being pleased to see you is not evidence that he is comfortable. The same paper notes that “acute stress can override pain responses”, which is why the consulting room is a poor place to judge chronic pain, and why owner-completed questionnaires exist.

So: dogs don’t conceal pain on purpose. Pain is missed because it arrives gradually, because ordinary sociable behaviour continues alongside it, and because a vet visit distorts the picture.

“Panting means pain”

Not supported by any validated instrument. Panting is not an item in the Glasgow scale, the Colorado State scale, the Canine Brief Pain Inventory, LOAD, the Helsinki index, or SNoRE. A targeted search for a primary study returned pet blogs and commercial sites only.

Cornell’s veterinary school does list excessive panting at rest as something to notice, so it is not baseless as a prompt, but treat it as a reason to look further, not as evidence. Heat, stress, cardiac and respiratory disease, and steroid treatment all cause panting.

“Fast breathing, racing heart and dilated pupils mean pain”

Contradicted directly:

“Physiological data such as tachycardia, tachypnea, systemic blood pressure, mydriasis… have proven to be ineffective as specific indicators of acute or chronic pain.” — Sander et al. 2025

“Read your dog’s face for pain”

There is no validated canine grimace scale. Two independent peer-reviewed reviews say so explicitly. A 2025 review: “Although there is no developed and validated grimace scale for dogs, similar to what was observed in other species, current behavioral-based pain scales include some aspects related to facial expression” (Frontiers in Veterinary Science 2025).

Validated grimace scales exist for horses, sheep, cattle, pigs, mice, rats, rabbits, and cats. Not dogs. A 2021 review explains why: canine facial anatomy varies enormously by breed, dogs have more than twenty facial movements, and humans reading dog faces bring “an anthropomorphic influence” (Mota-Rojas et al., Animals 2021).

Any infographic showing you a dog’s “pain face” is extrapolating from other species.

5. How much should you trust your own read?

This is uncomfortable, and you should have it anyway.

The caregiver placebo effect

In a randomised, double-blinded, placebo-controlled trial of 58 dogs with osteoarthritis lameness, researchers compared what people observed against force-plate measurement:

“A caregiver placebo effect for owners evaluating their dog’s lameness occurred 39.7% of the time. A caregiver placebo effect occurred 44.8% of the time when veterinarians examined dogs for lameness at a walk… This effect was significantly enhanced with time.” — Conzemius & Evans, JAVMA 2012

Roughly two owners in five saw improvement the force plate did not, and the veterinarians did it slightly more often. This is not a story about owners being unreliable. It is a story about human perception, and the professionals are in it too.

Note the last sentence especially: the effect grew over time. A confident “he’s so much better now” six weeks into a new medication is exactly where this bites hardest.

Subtle signs are where it breaks down

A 2026 study presented pain scenarios to 530 dog owners and 117 non-owners. For an obvious, movement-related case, owners did well. For a case describing subtle signs (shadowing family members, restlessness at night) the authors found “no differences… between dog owners and non-dog owners,” concluding “this may indicate that dog owners recognise subtle dog pain signs with less ease” (Gardeweg et al., PLoS ONE 2026).

Owning the dog conferred no advantage at all on the subtle case.

A companion study followed 51 dogs discharged after surgery, with owners logging behaviour daily and three veterinarians independently assessing video. It found “no correlation… between pain estimation by the veterinarians and the dog owners,” and concluded owners recognise changes but “lack standardized interpretation methods” and “emphasize movement-based changes over subtle behavioral shifts” (Gardeweg et al., PLoS ONE 2026).

So what does this actually mean for you

Not that your observations are worthless: the opposite. WSAVA is unambiguous: “Owner assessments are the mainstay of the assessment of chronic pain in dogs.” You are the only person who has seen this dog every day for a decade.

What it means is that instinct alone, unstructured and without a baseline, is demonstrably unreliable, for you and for your vet. The documented countermeasure is to record specific, concrete activities at intervals, rather than asking yourself the global question “is he in pain?”

Which is what the tools in the next section do.

Our note: We read the placebo finding as a problem with human perception, not owner competence. The two Gardeweg 2026 papers are recent.

6. Tools you can actually use

Most canine pain scales are for clinicians: the Glasgow scale requires kennel observation, leading the dog, and palpation. Two good ones are designed for owners and are free.

Canine Brief Pain Inventory: the one to start with

Owner-completed, validated in osteoarthritis and separately in bone cancer, and available as a free PDF from the University of Pennsylvania.

Eleven items in three parts: pain severity over the last seven days (at its worst, least, average, and right now, 0–10); how much pain interferes with general activity, enjoyment of life, rising from lying down, walking, running, and climbing stairs; and one overall quality-of-life rating.

→ Download the CBPI from Penn Vet · online version

Penn Vet states that the tools on that site “have been validated by investigators at Penn Vet and have sufficient science to support their use in the recommended populations.”

If your dog has cancer, this is the one. The CBPI was separately validated in dogs with bone cancer, which few instruments can claim.

Helsinki Chronic Pain Index

Also owner-completed, validated in osteoarthritis, and notably designed to be filled in “at home without prior instructions.” Eleven items covering mood, play, vocalising, walking, stairs, running, jumping, lying down, rising, and moving after rest and after exercise (Hielm-Björkman et al., AJVR 2009).

Two cautions. The 2009 paper reports no cut-off score: thresholds circulating online are not from it. And versions differ in total possible score, so check which you have.

LOAD: ask your vet for it

The Liverpool Osteoarthritis in Dogs instrument is owner-completed and well validated, but it is distributed under exclusive licence and the download is restricted to veterinary professionals. Your vet can give it to you; you cannot download it yourself.

What doesn’t exist

No validated dog-specific pain instrument exists for the hospice or terminal setting. We checked the end-of-life guidelines and a scoping review of quality-of-life tools. The field says so itself: the IAAHPC guidelines state there is “currently no ‘gold standard’ for assessing animal pain, and considerable research is still needed in this area.”

That is worth knowing when a quality-of-life scale is handed to you. A 2022 scoping review of generic quality-of-life tools for dogs and cats found that “although at least one aspect of reliability and validity was assessed for seven of the tools, none were validated across all measures” (Fulmer, Laven & Hill, Animals 2022).

Our note: Pointing you to the questionnaire first is our choice. Share the results with your vet.

7. How common is arthritis, really?

Three figures circulate and they do not agree, because they measure three different things. Rather than pick the convenient one:

Figure	What it actually measures	Source
2.5% annual period prevalence	Clinically diagnosed osteoarthritis, dogs of all ages, UK primary care, n=455,557	Anderson et al., Scientific Reports 2018
~40%	Detection via a screening questionnaire: traces through AAHA to a 2019 conference abstract, not a peer-reviewed study	AAHA 2022, refs 2–4
36–57% per joint	Radiographic change in dogs over 8, in a referral orthopaedic population	Roitner et al. 2024

Any article giving you a single “X% of dogs have arthritis” without saying which of these it means is being imprecise. Two other widely quoted figures, “20% of dogs over one year” and “80% of senior dogs”, we could not trace to a primary source at all.

The more useful numbers from Anderson 2018 are these: median age at first diagnosis was 10.5 years, and among diagnosed cases, 75% were recommended an analgesic, of which 77.9% actually received one. (Multiplying those gives roughly 58% of diagnosed dogs receiving pain relief, our arithmetic from their two figures, not a number the paper states.)

Diagnosis at 10.5 years, in a condition that develops over years, tells you something on its own.

8. When pain is not being controlled

There is no authoritative owner-facing “call the vet if…” list in the literature. What exists is better: thresholds built into validated instruments, and drug-safety guidance.

Score before you start. Complete the CBPI or Helsinki index before a new medication begins. Without a baseline, neither you nor your vet can distinguish real improvement from the caregiver placebo effect. On the CBPI, treatment success is conventionally a reduction of at least 1 in pain severity and at least 2 in pain interference.

Night-time restlessness returning is among the better-evidenced signals, per the trial in §3.

Reassessment intervals. AAHA guidance is that chronic pain patients “would then be reevaluated regularly (every 3–6 months).”

On anti-inflammatories, the FDA lists the signs to watch for: “Vomiting,” “Diarrhea,” “Bloody or black stool,” “Decreased appetite,” “Decreased activity level,” and “Yellowing of the whites of the eyes or the gums”, with the instruction to “stop giving the medication and call your veterinarian” (FDA). The FDA also advises asking about baseline blood tests before long-term use, and never adjusting a dose yourself.

Ask for your own trigger list. AAHA guidance says owners should receive “specific instructions, both verbally and in writing, including when the next assessment is recommended,” and be told what to do if adverse effects appear. If nobody has given you that, ask. Its absence is itself worth raising.

One hard thing about terminal pain

If your dog has cancer, this deserves saying plainly. A trial of dogs with bone cancer concluded:

“Dogs with bone cancer are affected by widespread somatosensory sensitivity characterized by peripheral and central sensitization and have a deficient inhibitory system. This severe pain is mostly refractory to palliative analgesic treatment.” — Monteiro et al., PLoS ONE 2018

This was a small open-label trial, thirteen dogs, so treat it as a signal rather than a settled finding. But an owner told “he’s on his pain medication, so he’s comfortable” is entitled to know that in bone cancer specifically, standard oral pain relief is often not sufficient, and that the pain involves mechanisms ordinary analgesia does not reach.

Our note: The Monteiro study is small, with 13 dogs. It is no reason to distrust your vet’s pain plan.

9. What to do this week

Write down five specific things he did easily a year ago. Stairs, the car, the sofa, greeting you at the door, sleeping through. Score each honestly now. Specific and concrete beats “is he in pain?”
Watch the nights. Restlessness between 11pm and 6am is the best-evidenced subtle sign here.
Complete the CBPI and take it to your vet. It takes five minutes and gives the conversation a shape.
Ask directly: “could he be in pain?” Not “is he slowing down?” The evidence says this conversation frequently doesn’t happen because each side waits for the other.
Ask what a trial of pain relief would look like. Sometimes the clearest answer to “is it pain or is it age?” is what changes when you treat it, with a baseline score recorded first, so the answer isn’t just what everyone hoped to see.

Sources

Guidelines

WSAVA Guidelines for Recognition, Assessment and Treatment of Pain, 2014. ⚠️ a 2022 edition exists (JSAP 2023;64(4):177–254) and was not read
2022 AAHA Pain Management Guidelines for Dogs and Cats
IAAHPC Animal Hospice and Palliative Care Guidelines. ⚠️ edition date unverified
2016 AAHA/IAAHPC End-of-Life Care Guidelines: notable for naming no pain-assessment instrument

“Just old age” and prevalence

Signs and mechanisms

Gruen, Samson & Lascelles. Sci Rep 2019;9:14192: the night-time activity trial
Mondino et al. SNoRE 3.0. Sci Rep 2023;13:13340
Mills et al. Animals 2020;10(2):318: pain and problem behaviour
Sander et al. Front Vet Sci 2025;12:1614403

No canine grimace scale

Owner assessment

Conzemius & Evans. JAVMA 2012;241(10):1314: caregiver placebo effect
Gardeweg et al. PLoS ONE 2026;21(4):e0344512
Gardeweg et al. PLoS ONE 2026;21(4):e0345418

Instruments

Terminal pain and drug safety

Always confirm with your vet before acting on anything on this page.