Is My Dog Depressed? Signs, Causes, Triage

A dog who has stopped greeting people at the door, stopped finishing meals, and started sleeping through most of the day is showing something real. The word owners reach for is depression. It is the wrong word, and the distinction is not pedantry. Dogs do not have a diagnosable depressive disorder in the way humans do. Veterinary behavioral medicine has validated diagnostic frameworks for canine anxiety disorders, noise phobias, and compulsive disorders — it has no such framework for depression, no diagnostic criteria, and no approved treatment for it as a condition. Anyone who tells you otherwise is describing something the field has not established.

Here is the accurate version. What owners call depression is a recognizable behavioral cluster: withdrawal, reduced engagement, appetite change, increased sleep, and loss of interest in activities the dog previously sought out. That cluster is real and it deserves to be taken seriously. It is also the least specific presentation in small-animal medicine. It is what a dog looks like with early kidney disease. It is what a dog looks like with an arthritic hip nobody has examined. It is what a dog looks like with a fractured carnassial tooth, with hypothyroidism, with a tick-borne infection, and it is also what a dog looks like six weeks after the household's other dog died. A "depressed" dog is a sick dog until a veterinarian says otherwise. This article covers what owners are actually observing, why the medical workup comes first and not second, which behavioral causes are genuine, what actually helps, and where medication does and does not belong.

What owners are actually seeing

The reports are remarkably consistent. Strip away the interpretive language and the same six observations appear:

• Reduced engagement. The dog no longer comes to the door, no longer follows the owner between rooms, no longer initiates contact. Solicitation behaviors — the nose nudge, the toy delivery, the paw — drop out.
• Loss of interest in previously reinforcing activities. A dog who reliably ran to the leash now stays on the bed. A dog who dismantled a food puzzle in four minutes now ignores it.
• Appetite change. Usually reduced — slower eating, leaving food, refusing kibble but accepting high-value items. Occasionally increased.
• Increased sleep, or increased time spent lying down while awake. The second one matters more than the first and is easy to miss.
• Social withdrawal. Choosing another room. Choosing under the bed. Moving away when approached rather than toward.

A dog who has always preferred a quiet sleeping spot may simply be resting. The guide to sitting alone in another room compares that baseline with sudden isolation, pain and illness signs.

• Reduced vocalization and reduced play, including play with other dogs the animal previously enjoyed.

Two of these can occur in a normal week. Four or more, sustained over several days, is a baseline change — and baseline change is the actual clinical signal, not any single behavior. A dog who has always been low-energy and quiet is not showing anything. A dog who was consistently forward and social and is now none of those things has changed, and the change is the data point.

The most useful thing an owner can do before any appointment is establish what the baseline actually was. Old phone videos are the best evidence available. Most people have footage of their dog from six months ago and have never watched it with clinical intent.

Why "depression" is the wrong word — and why the word matters

Human depressive disorders are diagnosed against explicit criteria that include reported subjective experience: worthlessness, hopelessness, recurrent thoughts about death. None of that is accessible in a dog, and no veterinary equivalent has been constructed and validated. Researchers can measure affective state in animals indirectly — judgment-bias paradigms, which test whether an animal interprets ambiguous cues optimistically or pessimistically, are a real and useful research method — but that is an experimental technique for measuring mood valence at a group level, not a clinical diagnosis for an individual dog in an exam room.

So the field does not say "your dog has depression." It says: here is a dog with reduced activity, reduced appetite, and social withdrawal, and here is the differential list.

The word matters because it steers behavior. "Depressed" is a psychological frame, and a psychological frame sends owners toward trainers, forums, calming chews, and a second dog for company. The correct frame is a physical exam. Every week that a dog with an undiagnosed painful joint or an untreated endocrine disorder spends being enriched, cheered up, and supplemented is a week of untreated disease.

The useful reframe is to stop asking "is my dog depressed?" and start asking "what changed, when did it change, and what has a veterinarian ruled out?" The second question has answers you can act on.

Sickness behavior: why medical always comes first

There is a specific biological reason this cluster is so unhelpfully generic.

Animals mount a coordinated behavioral response to illness and inflammation — reduced activity, reduced food intake, increased sleep, social withdrawal, reduced grooming, reduced exploration. Benjamin Hart described this as sickness behavior and argued it is an organized, adaptive strategy rather than a side effect of feeling unwell: the animal conserves energy and redirects resources toward immune function and fever. It is mediated by inflammatory signalling, and it is conserved across mammals.

The consequence for owners is direct. Sickness behavior and the "depression" cluster are, from the outside, the same set of outputs. Almost any inflammatory, infectious, painful, metabolic, endocrine, or neoplastic process can produce it. That is not a hedge — it is why the veterinary community treats acute-onset lethargy and withdrawal as a medical presentation by default.

The differentials that masquerade as low mood

This table maps what owners commonly report onto what a veterinarian is actually thinking about. It is a triage aid, not a diagnostic tool — most of these overlap heavily, and several can be present at once.

Notice the last row. Behavioral causes sit at the bottom of the list, and they are reached by exclusion. That ordering is the entire point of the article.

What a proper workup should include

Owners who arrive saying "my dog seems depressed" sometimes get a short appointment and a shrug. Arriving with structure changes that.

Bring three things:

1. A written timeline. When did each change start, in what order, and what else happened in the household that week. Dates, not "a while ago."
2. Video. Thirty seconds of the dog getting up from lying down, thirty seconds walking away from the camera, and thirty seconds at a meal. Gait and rising are where orthopedic pain shows itself, and they are exactly what a dog will not reproduce in an exam room.
3. Numbers. Daily food intake, water intake, hours of activity, and body weight if you can get it. Weight trend over months is one of the highest-value data points in the room.

What a reasonably complete first-pass workup looks like: full physical examination including orthopedic and neurologic assessment; a conscious oral exam with the explicit acknowledgement that it is limited; CBC, serum chemistry, and urinalysis; total T4 in dogs where the signalment fits; regionally appropriate infectious disease testing; blood pressure in seniors; and imaging where the exam points somewhere.

Two additions are worth asking about directly. A time-limited analgesic trial — a defined course of appropriate pain medication with a pre-agreed reassessment date — is a legitimate diagnostic tool when the orthopedic exam is equivocal, and chronic pain in dogs is very frequently equivocal on exam. Validated owner-report pain instruments exist and are underused: the Canine Brief Pain Inventory, the Liverpool Osteoarthritis in Dogs questionnaire, and the Helsinki Chronic Pain Index all give a structured score you can repeat in six weeks and compare. Ask for one.

For older dogs specifically, the overlap between pain, cognitive change, and sensory decline is dense enough that it deserves its own reading: see pain signals in older dogs, senior dog behavior changes, and cognitive dysfunction in senior dogs.

The behavioral causes that are genuine

Once the workup is clean, the behavioral list is real and worth working through properly.

Loss of a companion animal or a person. The most commonly reported trigger, and the one handled in its own section below.

Household composition change. A new baby, a partner moving in or out, a child leaving for university, a housemate's dog arriving. From the dog's perspective the change is not emotional in the way owners frame it — it is a wholesale rewrite of who reinforces what, where the resting places are, and when the reliable events happen.

Moving. New house, new smells, no established resting spot, different acoustics, different walk route. Reduced engagement in the first two to four weeks after a move is common and usually self-resolving if routine is rebuilt deliberately.

A schedule change, particularly the owner's. Return-to-office after a period of home working is a large and underrated cause. The dog's day goes from continuous low-level social contact and frequent short outings to eight hours of nothing. The visible result is not always distress — it is often flatness.

Reduced exercise and enrichment, usually for a hidden reason. The most common hidden reason is that the owner got injured, sick, busy, or older. Walks got shorter. The long-line decompression walk stopped. Nobody noticed the input dropped by half.

Chronic under-stimulation. Distinct from a recent drop. A dog who has never had adequate outlets can settle into a low-arousal, low-engagement default that owners read as a calm temperament until something reveals otherwise.

Aversive training and its aftermath. This one has evidence behind it. Vieira de Castro and colleagues (2020, PLOS ONE) compared companion dogs trained with aversive-based versus reward-based methods and found the aversive-trained dogs showed more stress-related behavior during training, higher post-training cortisol, and a more pessimistic profile in a judgment-bias task. A dog that has learned its behavior does not reliably change outcomes will do less behavior — the original learned-helplessness experiments of the 1960s were conducted on dogs, and while the modern neuroscientific reinterpretation is that passivity is the default state and control is what gets learned, the practical implication is unchanged. Suppression produces a quiet dog, and a quiet dog is not a well dog. See positive reinforcement vs balanced training and why dominance theory is wrong.

A recent adoption or rehoming. Newly homed dogs frequently show a period of flat, withdrawn, low-engagement behavior before their actual personality appears. The widely circulated "3-3-3 rule" that describes this in tidy intervals is folk guidance rather than validated science, and individual variation is enormous — but the underlying observation is sound: give a new dog weeks, keep the environment simple, and do not judge temperament in week one.

Sensory decline. Straddles medical and behavioral. A dog losing vision or hearing engages less because engagement has become harder and more startling, not because affect has changed.

When it's grief specifically

Dogs do show measurable behavior change after the death of a companion animal or person in the household. That much is not in dispute. What that change is remains genuinely open, and the honest answer is more interesting than the sentimental one.

The best-known evidence is owner-survey based. Uccheddu and colleagues (2022, Scientific Reports) surveyed owners of dogs who had lost a companion dog and reported clusters of change: increased attention-seeking, reduced play, reduced overall activity, more sleeping, reduced appetite, increased fearfulness, and increased vocalization. Those findings are useful and they are also limited in ways the authors acknowledged — retrospective owner questionnaires cannot separate the dog's response from the owner's own grief, from the household's altered routine, or from the owner's expectation of what a grieving dog should look like.

What is reasonably well supported: behavior changes after a household loss are common, they involve the same withdrawal-and-reduced-activity cluster this article is about, and they typically attenuate over weeks to a few months.

What is not established: that dogs understand death as a concept, that the response is emotionally equivalent to human bereavement, that dogs need to see the body, that there is a normal grieving period with a defined length, or that the response is distinguishable from a response to sudden social and routine disruption. The last point is the important one. When a companion animal dies, the surviving dog simultaneously loses a social partner, a play outlet, a sleeping arrangement, a feeding routine, and a predictable daily structure — while living with humans whose behavior has also abruptly changed. Untangling grief from disruption is not currently possible from the outside.

What to do practically:

• Hold the routine steady. Feed, walk, and settle at the same times. This is the single highest-value intervention and it is free.
• Change one thing at a time. Do not simultaneously remove the other dog's bed, rearrange the room, and start a new walk route.
• Do not rush a replacement. There is no good evidence that acquiring another dog helps a withdrawn dog, and it adds a substantial new stressor to a household already in flux. If another dog is right for the household, it will still be right in three months.
• Rebuild reinforcement density. The lost companion was providing engagement. Something has to replace that input — sniffing work, short training sessions, more one-to-one contact.
• Keep watching for medical signs. This is the safety point. Owners who have a grief explanation available tend to attribute everything to it, and a dog whose withdrawal is actually a developing illness gets weeks of sympathy instead of a blood panel. Attributing behavior change to grief is only safe after the exam is clean.

What actually helps

Assuming the medical workup is clean and the cause looks environmental or social, the interventions below are the ones with real mechanisms behind them. None of them are dramatic. All of them compound.

Predictable routine. Fixed feeding times, fixed walk times, a defined resting place. Predictability reduces the cognitive load of anticipating what happens next, and it is the foundation everything else sits on.

Sniffing work, before anything else. Scatter feeding in grass, snuffle mats, hidden-food searches indoors, and long-line decompression walks where the dog sets the pace and route. Olfaction dominates canine sensory processing, and sniffing work reliably lowers arousal per minute more effectively than the equivalent duration of physical exercise. For a flat, disengaged dog it also has a specific advantage: the effort threshold is low. A dog who will not chase a ball will often still put its nose down.

Graduated re-engagement. Start below the dog's current capacity, not at the level it used to manage. If the dog will not do a five-minute training session, do thirty seconds. Reinforce any voluntary participation — approaching, orienting, taking food, following one step. Stop while the dog is still engaged rather than pushing to the point of disengagement. The goal is to rebuild a history of "participating produces good outcomes," and that history is built in short repetitions, not long ones.

Food delivery as enrichment. Move meals out of the bowl and into puzzle feeders, scattered searches, or hand-delivered training reps. This raises daily behavioral output without requiring the dog to want anything new.

Exercise scaled to actual physical capacity. Critical caveat: this comes after pain has been ruled out. Pushing a painful dog into more walking makes things worse and is a common well-intentioned harm.

Sleep hygiene. Adult dogs sleep a great deal and need it undisturbed. A defined resting place away from household traffic, doorways, and children improves sleep quality — and poor-quality fragmented sleep produces exactly the daytime flatness owners are worried about.

Social contact matched to the individual. Some dogs re-engage faster with more human contact. Some need less pressure and more choice. Forcing interaction on a withdrawn dog — pulling it out from under the bed, insisting on petting, arranging a dog-park visit — reliably backfires.

Measure it. Pick four numbers and record them daily: grams eaten, minutes of voluntary activity, number of times the dog initiated contact, and hours slept. Review at two weeks. Owners are poor at detecting slow improvement without a record, and a record also gives the veterinarian something real if you go back.

Timeline expectation: a genuinely implemented plan should produce visible movement within two to three weeks. No movement in three weeks means the plan is wrong or the cause was never behavioral. Return to the veterinarian rather than continuing.

What people get wrong

"It's emotional, so I need a trainer." The most consequential error on this page. A behavior professional working a case that turns out to be orthopedic pain is running a protocol against a moving target, and good behavior professionals know this — a credentialed consultant will send you back to the veterinarian before opening the case.

"He's just getting old." Age is not a diagnosis. It is a risk factor that makes every item in the differentials table more likely, not less. The senior dog who "slowed down" is the single most under-investigated presentation in companion-animal practice, and osteoarthritis and dental disease are the two most common findings.

Reaching for supplements and CBD first. Evidence for cannabidiol in dogs is early and concentrated in osteoarthritis-associated pain, not in behavior; over-the-counter product quality and labelling accuracy are inconsistent; elevated liver enzyme values have been reported in dogs receiving it; and it can interact with other medications. Whatever its eventual role turns out to be, it is not a treatment for an undiagnosed behavior cluster, and buying it instead of an exam trades a diagnosis for a delay.

Getting a second dog to cheer the first one up. No good evidence supports this, the new dog is a significant stressor in itself, and if the original dog's problem is medical the household now has two problems.

"He's picking up on my mood, so there's nothing to do." Half true and badly applied. Sundman and colleagues (2019, Scientific Reports) found long-term stress synchronisation between dogs and their owners using hair cortisol — owner stress and dog stress track together over months. That is a real finding. It does not license inaction. It means the household's stress level is one of the variables worth changing, alongside everything else, and it certainly does not explain away a dog who stopped eating.

Reading a quiet dog as a well-behaved dog. A dog that has become still, undemanding, and easy is sometimes a dog that has stopped trying. Flatness is not calmness. The distinction is in whether the dog initiates — a calm dog still offers behavior when something interesting appears; a shut-down dog does not. More on reading the difference in what is my dog trying to tell me and signs your dog is stressed.

When medication is appropriate

Medication is a veterinary decision, full stop. Only licensed veterinarians can prescribe, and in this presentation the prescribing decision depends entirely on what the workup found.

There is no approved drug for canine depression, because it is not a diagnostic entity. What gets prescribed treats an identified condition. Levothyroxine treats hypothyroidism. Analgesics treat pain. Fluoxetine and clomipramine both carry FDA approval for canine separation anxiety, and both labels specify use alongside behavior modification rather than instead of it. Selegiline is approved for canine cognitive dysfunction. Each of those is a specific drug for a specific diagnosis — none of them is a treatment for "seeming sad."

Where psychoactive medication genuinely belongs is when the workup identifies an anxiety disorder, a compulsive disorder, cognitive dysfunction, or a pain condition with a behavioral component, and when a behavior modification plan is in place to be supported. Medication makes learning possible in an animal too aroused or too uncomfortable to learn. It does not do the teaching.

Two practical points. SSRIs typically need four to eight weeks at an adequate dose before efficacy can be judged, so a two-week trial tells you nothing. And any consideration of psychopharmacology is a strong argument for consulting a Diplomate of the American College of Veterinary Behaviorists (DACVB) — a veterinarian with a three-year residency in clinical behavior who can run the medical workup and manage the drugs in one place.

When to see a veterinarian, and who comes after

Same day, treat as an emergency: collapse or inability to stand; pale, white, or blue gums; laboured or open-mouth breathing at rest; unproductive retching with a distended or hard abdomen; straining to urinate with nothing produced; seizures; sudden severe pain; suspected toxin ingestion.

Within 24 to 48 hours: complete refusal of food beyond roughly a day in an adult dog — sooner in puppies, toy breeds, and diabetic dogs; refusal of water; vomiting or diarrhoea alongside the withdrawal; any measurable weight loss; withdrawal accompanied by any physical sign at all; a sudden change in a dog over about seven years old.

Within the week: gradual, isolated withdrawal in an otherwise physically normal adult dog with no other signs. This is the only version of the presentation where a routine appointment is reasonable, and it still needs an appointment.

Once the workup is clean, the professionals who handle the behavioral side are credentialed and identifiable. DACVB (Diplomate, American College of Veterinary Behaviorists) for anything with a medical or medication component. CAAB or ACAAB (Certified Applied Animal Behaviorist, certified through the Animal Behavior Society) for complex behavioral cases. CDBC (Certified Dog Behavior Consultant, through the IAABC) for fear, anxiety, and reactivity work. CSAT where separation-related distress is the specific problem — see dog separation anxiety.

Avoid anyone whose materials reference dominance, pack leadership, alpha status, or "balanced" methodology, and anyone who proposes to fix a withdrawn dog with corrections. The full vetting process is in how to find a credentialed behaviorist.

Building the record your vet actually needs

The hardest part of this presentation is that the evidence lives at home and the appointment happens somewhere else. Dogs mask in clinics. Gait normalises under adrenaline. The dog who has not risen from the bed in four days trots into the exam room looking fine.

PetTranslator.ai is built for the home half of that problem. Upload a clear photo of your dog and it returns a structured read of what is observable — posture, weight distribution, facial tension, ear and eye state — written the way a behaviorist writes a session note: markers first, interpretation second, a confidence score with its reasoning, and an explicit list of what could not be assessed from a single frame. Used across days, it gives you a dated record of change rather than a memory of it.

It does not diagnose, it cannot see a thyroid panel, and on this particular question it will point you at a veterinarian rather than at a training plan — because that is the correct answer. Its job here is to help you notice earlier and describe more precisely. The diagnosis belongs to your vet.

Sources

• Hart, B. L. (1988). "Biological basis of the behavior of sick animals." Neuroscience & Biobehavioral Reviews. The foundational account of sickness behavior as an organized adaptive response — the reason lethargy and withdrawal are treated as medical presentations.
• Overall, K. (2013). Manual of Clinical Behavioral Medicine for Dogs and Cats. Elsevier. The clinical reference for behavioral diagnosis and for the medical-workup-first principle.
• AVSAB Position Statement on Humane Dog Training (2021) — American Veterinary Society of Animal Behavior.
• Vieira de Castro, A. C., et al. (2020). "Does training method matter? Evidence for the negative impact of aversive-based methods on companion dog welfare." PLOS ONE. Stress markers and judgment-bias findings in aversively trained companion dogs.
• Uccheddu, S., et al. (2022). "Domestic dogs grieve over the loss of a conspecific." Scientific Reports. Owner-survey data on behavior change following the death of a companion dog — read alongside its own stated limitations.
• Sundman, A.-S., et al. (2019). Scientific Reports. Long-term stress synchronisation between dogs and their owners, measured via hair cortisol.
• AAHA Canine and Feline Behavior Management Guidelines — American Animal Hospital Association.
• AAHA Pain Management Guidelines for Dogs and Cats — American Animal Hospital Association. Includes guidance on recognising chronic pain and on validated owner-report pain instruments.
• American College of Veterinary Behaviorists — dacvb.org. Credential standards and diplomate directory.

Khabir Mughal is the founder of PetTranslator.ai. This article was reviewed against the AVSAB Position Statement on Humane Dog Training and Karen Overall's Manual of Clinical Behavioral Medicine before publication. It is not veterinary advice, and "depression" is used here only as the term owners search for — not as a veterinary diagnosis. For any sudden behavior change, see a veterinarian first.