Sundowning in Dogs: Why 3 A.M. Is the Hardest Hour

It’s somewhere between two and five in the morning, and your dog is up again. Pacing the hallway and standing in the corner facing the wall and panting at the foot of the bed, or barking at something you can’t see.

You’ve tried getting up with them, and you’ve tried lying still and hoping. Neither works, and you have to be functional in four hours.

Here’s the short answer. Sundowning in dogs isn’t a veterinary diagnosis. It’s a phrase borrowed from human dementia care, and what it describes in dogs is usually one of two things: canine cognitive dysfunction, the real condition behind most night-time confusion in old dogs, or something else entirely that’s treatable and worth finding first.

That second possibility is why this article starts where it does. A dog kept awake by arthritis, a full bladder, a failing heart or a drug started three weeks ago looks, at 3 a.m., almost exactly like a dog with dementia. That difference matters, because one of those has a fix.

What follows is what the research shows about the hours your dog is losing, what to rule out first, what helps, and what to do about the fact that you aren’t sleeping either. Take this information to your vet, but don’t use it as a substitute for one.

Key Takeaways

• Sundowning isn’t a veterinary diagnosis. It’s a human dementia term, and the 2026 consensus guidelines on canine cognitive dysfunction call the same thing sleep disruption instead.
• In a survey of 205 pet parents, over half of the dogs were most unsettled between 2 and 5 a.m., not at dusk.
• Canine cognitive dysfunction is a diagnosis of exclusion. Pain, high blood pressure, Cushing’s, sight and hearing loss, urinary and heart disease, and drug side effects all get ruled out first.
• You can count your dog’s sleeping breathing rate at home tonight. A rate above 30 breaths a minute, or one that keeps climbing, is worth a phone call.
• A dementia diagnosis isn’t a prognosis. In two long-term studies, dogs with cognitive dysfunction didn’t die sooner than dogs without it.

Why 3 A.M. Is the Hour, Not Sunset

Dogs with cognitive decline are most unsettled in the small hours, not at dusk. In a 2026 survey of 205 pet parents, over half reported their dogs were at their worst between 2 and 5 a.m. What separates affected dogs isn’t how long they sleep but how badly, and how much of that sleep slides into daylight.

The time you’re awake isn’t random, in other words, and it isn’t the hour the name predicts.

Dig into that survey, and the useful finding isn’t the timing. Trouble falling asleep, trouble staying asleep, and getting up to urinate in the night all tracked with how severe a dog’s decline was. Total hours slept didn’t.

There was a trend toward more late-afternoon restlessness as impairment progressed, so the evening isn’t nothing. It just isn’t where the worst of it lands.

That fits what happens when you wire a dog up. In a sleep-lab study of 28 senior dogs, the ones with higher dementia scores spent less time in both non-REM and REM sleep, with brain-wave patterns the authors described as shallower sleep. A Dog Aging Project analysis of more than 10,000 dogs aged over eight, published in GeroScience by Fisher and colleagues, found that dogs with cognitive dysfunction slept more during the day.

Put those together, and the 3 a.m. pacing stops looking mysterious. A dog who has slept in patches since lunchtime arrives at two in the morning with nothing left in the tank. They aren’t being difficult, and they aren’t reacting to the dark. They’re out of sleep, in a body whose clock has come loose.

Why Vets Don’t Use the Word Sundowning

Sundowning is a human dementia term, not a canine diagnosis. It doesn’t appear in the 2026 international consensus guidelines on canine cognitive dysfunction, which call the same sign sleep disruption instead, and it isn’t in Cornell’s client materials either. The word matters because it frames the problem as normal, which stops people from asking why.

The phrase comes from human medicine, where it describes late-day agitation in people with Alzheimer’s disease. It traveled into pet content because it’s evocative and sounds like it explains something.

In 2026, an international working group of twelve veterinary specialists published the first consensus guidelines for diagnosing canine cognitive dysfunction syndrome, and the sign they describe is sleep disruption, one of six domains, with no dusk effect attached to it. The word is likewise absent from Cornell’s client-facing page on cognitive dysfunction, which puts it plainly instead: wandering the house at night, sleeping more during the day.

Even in human medicine, the concept is shakier than you’d think. A 2016 review in Frontiers in Medicine noted that sundown syndrome doesn’t appear in the DSM-5 and has no standardized diagnostic criteria, with reported prevalence ranging from 2.5% to 66% depending on who was counting and how.

Dogs don’t run on a fully broken clock either. When one team put activity monitors on 33 aged dogs, they found no significant differences in daily activity rhythms between dogs with better and worse memory. The underlying rhythm held. What degraded was sleep quality, and how much of it got pushed into daylight.

So why does the word matter? Because a name that says, “this is the evening thing old dogs do,” closes the question.

One Australian survey estimated that 14.2% of dogs over eight met the criteria for cognitive dysfunction while only 1.9% had ever been diagnosed with it. That gap isn’t because vets can’t recognize it. It’s because the signs get filed under normal aging at home and never make it into the appointment.

Before You Call It Dementia: What Else Wakes an Old Dog

The 2026 consensus guidelines are blunt about this. A cognitive dysfunction diagnosis requires identifying alternate causes through physical, orthopedic, and neurological examination and laboratory work, and it requires that the signs persist after any other conditions found have been treated.

In plain terms: dementia is what’s left when the other things have been ruled out, not the first thing you land on. Go slowly here. Of everything below, this is the part most likely to change something in your house.

Pain, Which Is the One Most Often Missed

Arthritis is the obvious candidate and the one most reliably overlooked, and that isn’t a failure of attention on anyone’s part. It arrives so gradually that it looks like ordinary aging, and that camouflage is the whole problem.

When UK researchers interviewed pet parents about the run-up to their dogs’ arthritis diagnosis, the changes they reported were subtle and intermittent, and had been going on long before anyone named the cause; few had connected them to arthritis at the time. A related survey found the mechanism behind it: people judged a sign as less urgent if they’d seen it before. If you’ve spent months telling yourself it was just age, you were doing what nearly everyone does with a slow change.

Be careful with what you’ll find written elsewhere, though. No good evidence shows that dogs’ arthritis pain follows a daily rhythm that peaks at night. The likelier explanation is duller and more fixable: nothing is distracting them, they’ve stiffened up after lying still, and the floor is hard.

When one team tested this properly, anti-inflammatories didn’t change how much dogs moved at night, though their families reported better sleep quality. That’s the split you’d expect if comfort improved without the movement changing.

Look for the tells that aren’t limping: reluctance to settle, circling before lying down, shifting position all night, groaning on getting up, choosing the cold hard floor over the bed, or a dog who no longer takes the stairs. We’ve written about the pain signals dogs actually give, most of which aren’t vocal.

Anything That Makes a Dog Need to Get Up

A dog who wakes because their bladder is full isn’t a confused dog, but at 3 a.m. the two look identical.

Cushing’s disease is the classic mimic, and the giveaway is the company it keeps: drinking and urinating far more, a ravenous appetite, a pot belly, thinning skin, muscle wasting, and panting. Diabetes produces the same thirst and the same night trips, and often cataracts on top, so the dog is newly half blind as well.

Then the plumbing itself. A urinary tract infection causes urgency and repeated fruitless trips to the door, and kidney disease means they physically have to get up. Both are found with a urine sample and a blood panel, and both are treatable or manageable.

High blood pressure deserves its own line because it’s silent until it isn’t. The American College of Veterinary Internal Medicine consensus statement lists sudden blindness, retinal detachment, bleeding in the eye, and neurological signs among the damage untreated hypertension causes. It reports that high blood pressure is common in dogs with Cushing’s. It’s measured with a cuff in a couple of minutes, and it’s worth asking for by name.

Ears and Eyes

Hearing loss is the mimic nobody expects. When researchers formally hearing-tested 39 elderly dogs, every dog in the most hearing-impaired group scored as abnormal on the dementia scale, and worse hearing went with higher dementia scores even after accounting for other factors.

Deafness and dementia get tangled together, and they’re easy to confuse, because a dog who can no longer hear you moving around the house at night has genuinely lost their bearings. Deafness is manageable even when it can’t be reversed. Wake a deaf dog by putting a hand flat on the bedding so they feel the vibration first, or let them catch your scent, rather than touching them out of nowhere.

Sudden blindness does the same thing faster. A dog who goes blind over a few days will bump into things, get stuck in corners and vocalize, and none of that has anything to do with their memory. Keep the furniture exactly where it is, keep their routes clear, and use texture underfoot to mark the places that matter.

Breathing and the Heart

Nothing else on this list has a shorter fuse, and you can run a free test tonight.

Wait until your dog is genuinely asleep, count their breaths for thirty seconds, and double it. In a study of dogs whose heart failure was well controlled on medication, the median sleeping breathing rate was 20 breaths a minute, and only one dog in fifty-one was at 30 or above. Dogs with early, still-silent heart disease sit under 30 too, and the day-to-day variation is small, which is why cardiologists treat 30 as the line worth watching. A sleeping rate persistently above it, or a number that climbs week on week, is worth a phone call rather than a wait.

The other distinction is simpler than it sounds. You can usually settle a confused dog if you go to them. A dog who can’t breathe can’t settle at all.

If they won’t lie flat, stand with their elbows held away from their body and their neck stretched out, or their gums look pale, gray or blue, that’s tonight rather than tomorrow. Our guide to the signs of congestive heart failure covers what else to watch for.

The Medicine Cabinet

First, write out every drug and supplement your dog takes and the date each one started. Then look at when the nights changed.

Steroids are the highest-yield suspect. Panting, restlessness, drinking and urinating more are all recognized effects of steroids, so a dog who started prednisone in April and stopped sleeping in May has told you something. Never stop steroids abruptly, but a conversation about dose, and about giving it in the morning rather than the evening, is a reasonable one to have.

Furosemide, given for heart failure, causes night urination by design. Ask the prescribing vet whether the timing of the evening dose can move, and don’t change it yourself, because that schedule is doing a job.

Gabapentin and trazodone are prescribed to help with exactly this problem and occasionally do the reverse, causing agitation, or leaving an already wobbly dog too unsteady to feel safe on their feet. Timing is the diagnostic tool here, and you’re the only one who has that information.

The Red Flags That Mean Tonight, Not Next Week

Most night-time restlessness isn’t an emergency. That’s worth saying before the list, because it reads frighteningly at 4 a.m., and most dogs who potter, get up, and settle again when you guide them back to bed need an appointment rather than an ambulance.

These are the ones that don’t wait, adapted for the sleepless-night reader, based on Cornell’s guidance on when to go to the emergency room.

Go Tonight

Any one of these on its own is enough. You don’t need a second sign to justify the trip, and you don’t need to wait to see whether it settles.

• Breathing that’s fast or labored at rest, refusing to lie flat, neck extended, elbows held out from the body, or the belly heaving with each breath.
• Gums that are pale, white, gray, blue, or brick red.
• Collapse, or a sudden inability to stand or use the legs.
• A swollen, drum-tight belly with retching that brings nothing up.
• Straining to urinate and producing nothing.
• A first-ever seizure, one lasting more than about five minutes, or more than one in a day.
• Repeated vomiting or diarrhea across a day, particularly alongside lethargy or known kidney or liver disease.
• Distress you can’t settle at all: crying out, rigid, trembling, unable to lie down. That’s what uncontrolled pain looks like.
• Sudden blindness, hugely dilated pupils, or blood visible in the eye.
• A head tilt, circling, or one-sided weakness that came on suddenly, especially with vomiting or an inability to stand.

Go This Week, and Don’t Let It Drift

None of these needs an overnight clinic. All of them get worse or harder to treat if they sit for a month, which is what usually happens when a change gets filed under old age.

• Night restlessness that started within a few weeks of a new medication or a dose change.
• New or worsening drinking and urinating.
• A sleeping breathing rate that’s trending upward, or sitting above about 30.
• A gradual, creeping head tilt, or circling that always goes the same direction. Cognitive dysfunction is symmetrical, so one-sided signs point somewhere else.
• Yelping when touched, lifted, or getting up.
• Weight loss, or a lump that’s changing.

If you’re stuck at 2 a.m., genuinely unsure which side of the line you’re on, our guide to deciding between the emergency vet and waiting until morning is built for exactly that call.

How Cognitive Dysfunction Is Actually Diagnosed

No blood test diagnoses canine cognitive dysfunction. Vets diagnose it by exclusion: a consistent history of progressive signs across six domains, other causes ruled out by examination and lab work, a normal or symmetrically abnormal neurological exam, and signs that persist after treating any other conditions found. An MRI is used only when warranted.

That’s why this needs a real appointment, not a quick word at the end of a vaccination visit.

Those six domains are worth knowing before you go, because they’re what you’ll be asked about: disorientation, changes in how your dog interacts with you, sleep disruption, house soiling and lost learned behavior, changes in activity, and anxiety.

That framework is scored with a free tool called the DISHAA assessment, a downloadable questionnaire that covers all six domains and can be filled out at the kitchen table. A pet food company distributes it, which is worth knowing, though the domains themselves were adopted by international consensus.

Bring two things and the appointment goes better. A written list of every medication and supplement with start dates, and video. Thirty seconds of your phone pointed at the hallway at 3 a.m. tells a vet more than ten minutes of your best description, and it captures the thing that never happens in the consulting room.

If this is your first time navigating an aging dog’s health as a whole, our senior pet care section covers the wider ground.

What Actually Helps at Night

The things that cost nothing have the least research behind them, and they’re still the best place to start because they’re free, safe, and address what’s actually broken. What pet parents tell us helped was rarely the supplement. It was the rug on the tile and the light left on in the hall.

Almost none of the environmental advice below has a controlled trial behind it, and nobody has ever studied night lights for dogs. Take it as what it is: plausible, harmless, and worth two weeks of your effort.

• Keep the schedule identical, especially the last meal, the last walk, and lights out.
• Leave a low light on where they sleep, so waking isn’t waking into blackness.
• Any hard floor they cross at night needs a rug or a runner. A dog who slips once will hesitate for weeks.
• Block the gaps behind and beside furniture, where a dog who can no longer reverse gets stuck.
• Water should be somewhere they can reach without having to navigate to it.

Two more things are worth trying because they’re cheap and the failure mode is nothing.

The first is a small snack at bedtime. If your dog wakes in the small hours eating grass, lip-smacking, swallowing repeatedly or bringing up yellow bile, that pattern is bilious vomiting syndrome, which happens when the stomach has been empty too long, and a late meal sometimes settles it entirely. If the vomiting carries on regardless, stop experimenting and get it looked at, because the same picture fits several less benign things.

Proximity is the other one. Moving their bed nearer to you, or yours nearer to them, resolves a surprising share of night vocalizing, because for many dogs the problem is being alone and unable to work out where you went.

Fix the Daytime Before You Fix the Night

Daytime is the bigger lever, and it’s the one people skip. If your dog sleeps from ten to four every day, there’s no sleep left to spend after midnight. Getting them up, out, sniffing, and mildly worked in daylight is the closest thing to an intervention here.

A very large Dog Aging Project analysis found less active dogs had substantially higher odds of cognitive dysfunction. The authors are careful to say the design can’t establish cause. Cognitive decline making dogs less active remains a live explanation, though their own sensitivity analyses lean the other way slightly. Take it as a good reason to try, not as proof it prevents anything.

Our roundup of comfort tools for anxious senior pets covers the physical props worth having.

What the Medicines Can and Can’t Do

One drug is FDA-approved for canine cognitive dysfunction, and it controls signs rather than slowing the disease. Diets and supplements have modest, manufacturer-funded evidence behind them. Nothing available today reverses the condition, and the most useful thing most pet parents can do is change the environment rather than reach for the medicine cabinet.

The honest version of this field is far less impressive than the marketing around it.

Selegiline, the Only Approved Drug

Selegiline, sold as Anipryl, is the only drug the FDA has approved for canine cognitive dysfunction, cleared for that use in December 1998. It’s worth knowing what the approval actually says. The indication is control of clinical signs, not slowing the disease, and there’s no evidence in dogs that it changes the underlying decline, regardless of what you read elsewhere.

In the placebo-controlled portion of the FDA trial, roughly half the dogs on the full dose improved on sleep and on activity at four weeks, against a substantial placebo response. Give it a full month or two before judging it. And note the sting in the tail: the same FDA summary records restlessness, irritability or hyperactivity in 15% of treated dogs against 6% on placebo, so in a minority it makes the night worse.

Interactions matter more than efficacy. Selegiline shouldn’t be combined with antidepressants like fluoxetine or clomipramine, with tramadol, or with flea and tick products containing amitraz, because of the risk of serotonin syndrome. Coming off fluoxetine means waiting at least five weeks before starting it, and going the other way means at least two.

Plenty of senior dogs are already on something from that list. This is a conversation with the prescribing vet who holds the full medication list, not something to add.

Diets and Supplements

On diet, one point cuts through much of the confusion. The randomized trial in dogs with diagnosed cognitive dysfunction tested a therapeutic diet containing medium-chain triglycerides plus a blend of B vitamins, antioxidants, and fatty acids, not the over-the-counter senior food with a similar-sounding name.

The manufacturer funded it entirely, and the control group improved in four of six domains, which suggests how much of this is expectation. It’s a reasonable, low-risk thing to try. It isn’t proven to change the course.

Supplements are thinner still. A 2025 systematic review of 30 studies found a median of 11.5 dogs per test group and a power calculation in only 11% of the trials. Omega-3s at a real dose and SAMe have the least-bad evidence. Most of the rest is hope with a label.

The most telling figure in this whole area comes from a 2025 survey of 318 American vets: nearly nine in ten recommend supplements, yet around 90% rated current strategies only slightly or moderately effective. Your vet’s uncertainty here isn’t a failure on your vet’s part. It’s the state of the evidence.

Three Safety Notes That Matter More Than Any of the Above

Never give your dog human melatonin without reading the full ingredient list. Sugar-free, fast-dissolve, and gummy sleep aids can contain xylitol, which causes a rapid and dangerous drop in a dog’s blood sugar, and which the Merck Veterinary Manual links to acute liver injury at higher doses.

A second problem exists even when it’s xylitol-free. When researchers analyzed 25 melatonin gummies sold in the US, 22 of them were inaccurately labeled, with actual content ranging from 74% to 347% of the stated dose. You can’t dose a dog reliably from a bottle like that. And melatonin has no controlled trial behind it in dogs for sleep at all. It’s used because it’s cheap and generally well tolerated, which is a fair reason but not the same as evidence.

The same warning applies to liquid gabapentin. Some human oral solutions are sweetened with xylitol, so if gabapentin is prescribed, ask specifically for capsules or a xylitol-free compounded liquid.

Last, the lavender diffuser. The American Society for the Prevention of Cruelty to Animals lists lavender as toxic to dogs, and the evidence that it calms them rests on three small studies that measured how much dogs moved rather than how anxious they were.

If you diffuse anything, the Merck Veterinary Manual says to keep pets out of the room entirely, run a passive diffuser rather than an ultrasonic one for under thirty minutes, and ventilate afterward. Never tea tree, pennyroyal, wintergreen, eucalyptus, cinnamon, or pine near a dog at all, and if there’s a cat or a bird in the house, don’t diffuse.

What a Dementia Diagnosis Does and Doesn’t Tell You

A diagnosis of cognitive dysfunction isn’t a prognosis. In two Danish cohort studies that followed older dogs for years, dogs with the condition did not die sooner than dogs without it, and the rate of decline varied enormously between individuals. What changes over time is how much support a dog needs, not how much time is left.

That surprises almost everyone, including people who work in this field.

Two Danish cohort studies tracked older dogs for years. The first followed 94 dogs over eight years old from 2008 to 2012 and found four signs dominating in the dogs with cognitive dysfunction: sleeping during the day and restlessness at night, less interaction, disorientation at home, and anxiety. Your night is on that list.

The key finding, though, is what happened next. Cognitive dysfunction did not shorten survival, and the authors wrote that dogs with it seem to have a good chance of living a full lifespan if supported by their vet and their family.

A second study following 51 dogs over two years reached the same conclusion on survival, and found that 33% of cognitively normal dogs slipped into mild impairment over that period while 22% of the mildly impaired progressed further. Decline rates varied enormously between individual dogs.

So a diagnosis isn’t a countdown. What tends to change over time isn’t how long you have; it’s how much support your dog needs and how much your household can carry. Those are two different questions, and it helps to keep them apart.

The first of those two studies flagged one more thing worth acting on. Its authors singled out anxiety as deserving special attention, because it’s stressful for the dog and for everyone else. If your dog is frightened at night rather than merely awake, say that out loud at the appointment. It’s a treatable target in its own right, separate from the cognition.

The Toll on the Person Doing the Caring

Every article on this subject has a paragraph telling you to practice self-care, and every one of them is useless, because you already know and it isn’t a knowledge problem. So here’s the research instead, which is more respectful of what you’re actually doing.

A clinical neuropsychologist at Kent State, Mary Beth Spitznagel, brought the concept of caregiver burden across from human dementia research and started measuring it in people caring for sick animals.

When her team compared people caring for a seriously ill pet with people caring for a family member with dementia, the human caregivers carried more overall burden, as you’d expect. But on three specific items the two groups were statistically indistinguishable: fear of the future, guilt, and worry about money.

You aren’t being dramatic. That’s measured.

What we hear from people at this stage is rarely about the dog. It’s some version of, “I don’t know how much longer I can do this,” usually followed by an apology for saying it.

In a survey of pet parents caring for dogs with cognitive dysfunction specifically, 16% were carrying a clinically significant burden of care, and one of the factors linked to it was simply where in the house the dog slept. Small logistics, large consequences.

What Actually Lightens It

One finding is genuinely encouraging. When researchers ran five weeks of group training classes for dogs with mild to moderate cognitive dysfunction, the dementia scores didn’t budge, but the caregivers’ burden scores fell anyway, in both study arms, along with reports of more confidence and more social support.

Nothing about the dogs’ brains changed. The weight of it got lighter. If you take one thing from this section, take that: the load is modifiable even when the disease isn’t.

Practically, split the nights with someone if there’s anyone to split them with, and take the shift that lets you get one unbroken stretch. Stop treating a closed bedroom door or a night in the spare room as a betrayal. Sleeping is maintenance on the person doing the caring, and our piece on the emotional toll of long-term pet caregiving goes further into what actually shifts it.

Weighing Quality of Life in a Dog With Dementia

There’s no threshold in the veterinary literature at which a dog with cognitive dysfunction should be euthanized, and this section takes no position on what anyone should do. What it offers is structure: free, credible quality-of-life scales, scored the same way each week, so that a decision rests on a trend rather than on one bad night.

Some people reading this are nowhere near that question. Some are sitting inside it tonight. Both are ordinary.

Start with the honest picture. In a UK study of nearly 30,000 dogs with a recorded method of death, 91.5% were euthanized, and among those, the highest odds of euthanasia rather than unassisted death attached not to specific diseases but to a judgment of poor quality of life and to behavior a household couldn’t manage. That’s population context, not a prediction about your dog.

What most of those decisions lack is structure. In a 2026 survey of 228 pet parents across seventeen countries, only 7.3% used a structured quality-of-life tool. Most people were going on personal observation, at the worst possible moment, from memory.

Use a Scale, Not Your Memory of Last Night

You can do better than memory, and it costs nothing. The JOURNEYS quality-of-life scale is free and fits this situation unusually well, because two of its eight categories are the ones most scales skip: uncertainty and understanding, which is where confusion lives, and a category called “You,” which scores whether the person doing the caring can keep doing it.

The Ohio State University’s veterinary hospital publishes a free assessment with one line in it worth the whole download: sometimes a single item, pain being the obvious one, indicates poor quality of life even when everything else still looks fine. Don’t chase a total.

Score the same tool on the same day each week and keep the sheets. Look for the direction of travel, not the number.

A written record protects you from the two distortions that hurt most: judging everything by last night, and failing to notice a slow slide because you adapted to it a week at a time. If a formal scale feels like too much, our good days and bad days tracker does the same job with a calendar and a pen.

One more thing, said plainly. Anyone quoting you a fixed threshold for this decision has invented it.

Researchers have found that how burdened a caregiver feels tracks with movement toward that decision even after accounting for how the animal is actually doing. That’s a correlation in a cross-sectional study, and it isn’t an accusation. It’s a reason to get support and veterinary input to lower the burden first, so that whoever has to decide is rested and informed rather than deciding after a fourth broken night.

What to Do Next About Sundowning in Dogs

Stop asking whether it’s sundowning and start asking what’s waking your dog. That means a blood pressure reading, a urine sample, a look at their joints, a hard review of every medication and its start date, and thirty seconds of video from 3 a.m. taken to a vet with the time to watch it.

“Is this sundowning?” is the wrong question because it’s a borrowed word that closes the subject. The better one has a real answer often enough to be worth chasing properly.

If the answer turns out to be cognitive dysfunction, that’s a diagnosis you can work with. Light in the hallway, rugs on the tile, a snack at bedtime, their bed moved nearer yours. The same day in the same order, every day, and something to do while it’s light out.

And if they’re frightened rather than just awake, say so, because that part is treatable. For everything else that comes with this stage, our dog end-of-life section covers the rest.

The nights are the price of the years, and the years were not nothing.