We Measured Dog REM Sleep and Its Impact on Behavior

What is happening when your dog twitches in sleep?

Ordinary twitching, facial movement, soft vocalization, and changes in breathing can occur during canine REM sleep, but home observation cannot confirm a sleep stage or reveal dream content. M. A. Crist, a clinical assistant professor at Texas A&M’s College of Veterinary Medicine and Biomedical Sciences, describes these movements as possible features of REM sleep while emphasizing that people cannot know directly whether a pet is dreaming or what the dream contains. Texas A&M’s veterinary explanation of canine REM behavior provides useful context, but it is not a diagnostic rule.

The twitching happens because REM sleep combines an active brain with reduced muscle tone. Muscle tone means the small amount of ongoing tension that helps the body hold posture and move. During REM, the brain may be highly active while most large voluntary muscles are inhibited. Small movements can still appear, especially around the eyes, muzzle, paws, and tail.

That pattern resembles the broad human description of REM sleep, but a visible movement is only an observation. A dog may twitch while shifting position, responding to a sound, or passing through another sleep state. A camera can document timing and frequency, but it cannot identify REM without physiological measurements such as electroencephalography.

Electroencephalography, or EEG, records electrical activity from the brain. In canine sleep studies, researchers combine EEG with eye-movement recordings, muscle activity, heart rate, breathing, and behavioral observation. Those combined measurements allow researchers to distinguish wakefulness, drowsiness, NREM sleep, and REM sleep more reliably than an owner can by watching from the sofa.

This distinction matters because owners often make one of two mistakes:

• They assume every twitch is a dream and become unnecessarily worried.
• They dismiss repeated, forceful, or unusual episodes as normal dreaming.

A brief paw flick followed by peaceful sleep is usually an observation to record, not a crisis. Repeated violent movements, prolonged stiffening, injury, confusion after waking, breathing difficulty, or episodes that resemble seizures deserve veterinary attention.

For a broader look at posture, nighttime restlessness, and changes in how a dog settles, see this guide to dog sleeping positions and joint health. A sleeping position is one piece of context, not a diagnosis of pain, anxiety, or sleep quality. You can also compare the broader relationship between comfort, placement, and rest in this discussion of how the right pet bed supports sleep and health.

The canine REM sleep cycle has two main phases

Dogs cycle through NREM and REM sleep rather than remaining in one uniform state. NREM generally provides quieter, deeper physical rest, while REM features more active brain rhythms and the movements owners associate with dreaming. Both phases matter, and a healthy sleep pattern depends on the sequence and continuity of the whole cycle.

NREM sleep: the quieter foundation

NREM sleep is often described as slow-wave sleep or deep sleep when the brain shows slower electrical patterns. The body tends to be still, breathing is more regular, and muscle activity is easier to distinguish from the small movements often seen in REM.

NREM is not an inactive period. Canine EEG research has identified 9–16 Hz electrical transients during NREM sleep that function as proposed spindle analogues. Sleep spindles are brief bursts of rhythmic brain activity associated with information processing in several studied species. In dogs, these NREM transients have been associated with learning and recall performance, but the research does not establish that they are identical to human sleep spindles or that REM is the only sleep stage involved in memory consolidation. The canine NREM EEG study on sigma-range transients supports a role for NREM activity in learning-related processes.

This is one reason the phrase “deep sleep in dogs” should not be treated as a synonym for REM. Deep NREM sleep and REM sleep have different physiological signatures and may contribute to different parts of recovery and learning.

REM sleep: active brain, quieter large muscles

REM sleep is marked by brain activity that becomes more similar to wakefulness than the slower patterns seen in deep NREM. Eye movements occur beneath the eyelids, breathing can become less regular, and small muscle movements may appear.

A 1997 EEG study of three trained laboratory dogs reported mean slow-wave sleep bouts of 2.2–4.4 minutes, mean REM bouts of 3.5–4.6 minutes, and REM latency of 33.9–41.8 minutes. REM latency is the time between sleep onset and the first observed REM period. These values describe a very small laboratory sample recorded during daytime sleep; they are not a universal target for household dogs. The early canine EEG recording study indexed in PubMed shows why measured REM periods can be short and repeated rather than one long block.

The brain activity seen in REM has attracted attention because it may support emotional and memory-related processing. Still, canine evidence remains more cautious than popular explanations suggest. Research has found EEG changes after learning and relationships between REM EEG features and later performance, but no checked canine study directly recorded a dog replaying a specific event from the day.

NREM and REM compared through a practical stability lens

The requested Canine REM Stability Index, or CRSI, is not a validated veterinary metric. No checked study supplies a formula, normal range, cross-breed benchmark, or clinical threshold. The table below uses “CRSI lens” as a practical way to organize observations, not as a score you can calculate or use to diagnose a dog.

The most useful home measure is therefore not “How many REM minutes did I see?” It is whether your dog can settle, remain asleep, wake normally, and function normally during the day.

How do REM brainwaves relate to learning and emotion?

Canine EEG studies suggest that sleep after learning is associated with measurable changes in brain activity and later task performance, but they do not prove that REM alone stores memories or creates a calmer temperament. The evidence supports sleep as part of learning and emotional processing, while leaving the exact role of each sleep stage open.

In one study, adult pet dogs learned unfamiliar command-word associations and then underwent three hours of polysomnography. Researchers found changes in NREM and REM EEG spectra after learning. Among dogs with usable REM data, post-sleep performance was associated with lower REM delta power and higher REM beta power. The reported relationships were correlations, not proof that REM caused the improvement. The task also measured a specific command-learning outcome rather than broad household behavior. The EEG and command-learning study by Kis and colleagues explains the finding and its limits.

A useful analogy is to think of sleep as a sorting period rather than a single filing cabinet. NREM may help stabilize some forms of newly learned information. REM may participate in other kinds of brain activity, including associations between emotional or sensory information. The analogy helps explain why researchers study both phases, but it should not be mistaken for direct evidence of a dog replaying its day.

The emotion evidence is similarly specific. In a within-subject experiment involving 15 adult family dogs, researchers disturbed REM sleep in one condition and interrupted NREM sleep in another before testing how the dogs responded to audiovisual displays of human emotion. Sleep disturbance affected performance on that task. Yet the REM-deprivation condition involved more awakenings, both conditions disturbed sleep, and the results were linked more clearly with relative NREM duration and the number of awakenings than with REM duration alone. The canine REM-versus-NREM emotion-processing experiment does not establish that uninterrupted REM prevents aggression, anxiety, or irritability.

That distinction is important for owners. If a dog sleeps poorly and behaves differently the next day, sleep may be part of the explanation. Pain, gastrointestinal illness, noise, fear, medication effects, age-related cognitive change, and household stress can also disturb sleep and behavior. A sleep problem may be a sign of another problem rather than the sole cause.

Puppies, adults, and seniors do not share one REM pattern

Age changes canine sleep electrophysiology, so a puppy should not be judged against an adult dog’s pattern. Breed and body size may also influence development, but the available research does not support a deterministic ranking of breeds by REM stability.

A developmental study of family dogs found age-related changes in sleep macrostructure and EEG power. In a juvenile sample of 60 dogs, time spent in REM was negatively associated with age between approximately 2 and 6 months. An extended sample included 91 dogs ranging from 2 to 30 months. These are developmental associations, not diagnostic thresholds. The developmental sleep electrophysiology research by Reicher and colleagues does not show that every puppy should spend a fixed percentage of sleep in REM.

Puppies may appear to twitch, vocalize, or move more often simply because their sleep architecture is changing rapidly. That does not make every unusual episode harmless, but it does make age a necessary part of the interpretation. A young dog who occasionally paddles, squeaks, and then wakes normally is different from a dog who has escalating, forceful, or injurious episodes.

Adult dogs also vary according to what happened before sleep. In a study of 16 family dogs, an active day was associated with more total sleep and more NREM and REM sleep. Nighttime recordings also contained more NREM and REM sleep than daytime recordings. The study included physically and mentally demanding activity, so its results should not be turned into a promise that exhausting a dog will automatically produce better rest. The study of pre-sleep activity and canine sleep electrophysiology supports the more modest conclusion that daily activity, timing, and location can influence sleep measurements.

Senior dogs require another layer of care. In a polysomnographic study of 28 dogs aged 10.4–16.2 years, higher canine dementia scores and poorer problem-solving performance were associated with less NREM and REM sleep during an afternoon recording. The design was observational, the recording covered an afternoon nap, and age-related disease could not be separated cleanly from sleep changes. The senior-dog sleep and cognition study supports veterinary attention to changes in sleep and cognition, not a claim that adding REM reverses cognitive decline.

Why can interrupted sleep change daytime behavior?

Repeatedly disturbed sleep can coincide with poorer daytime behavior and learning, but the canine evidence does not establish one simple REM-to-aggression pathway or prove that every interruption raises cortisol. The safest conclusion is that sleep continuity is worth protecting while the underlying cause of poor rest is investigated.

A large caregiver questionnaire study of 1,330 dogs found associations between shorter reported sleep, greater ease of disturbance, and greater caregiver-reported problem-behavior severity. The study did not measure REM or EEG, did not identify an optimal number of sleep hours, and could not establish whether poor sleep caused behavior changes or whether another household or medical factor affected both. The caregiver-reported canine sleep and behavior study is useful for identifying a pattern, not for calculating a treatment threshold.

The requested Behavioral Sleep Quality Ratio, or BSQR, is also not a validated canine measure. No checked study defines its numerator, denominator, cutoff, reliability, or predictive value. You may use the idea as an observation framework:

• Sleep continuity: Does your dog settle and stay asleep, or wake repeatedly?
• Daytime function: Is your dog alert, responsive, and able to perform familiar routines?
• Stress signals: Are pacing, repetitive behavior, startle responses, or avoidance increasing?
• Context: Did pain, noise, heat, medication, visitors, travel, or schedule changes occur?
• Persistence: Is the pattern present for one night, several nights, or many weeks?

This is more useful than assigning an invented number. It keeps sleep observations connected to real behavior without giving the appearance of a clinical score.

The hormone question deserves similar care. The supplied canine sleep-interruption studies measured EEG, sleep stages, awakenings, and behavior; they did not establish that REM disruption raises cortisol or that a cortisol increase delays learning through a demonstrated canine pathway. Human and rodent findings may suggest biological possibilities, but they cannot be presented as direct canine evidence.

Pre-sleep experience can affect sleep architecture in surprising ways. In one study of 16 adult pet dogs, positive and negative social treatments produced different sleep latency and stage distributions. Relative REM was lower after the positive treatment in that experimental setting. That result is a useful warning against simplistic claims that “more REM always means better emotional health.” The study of social experience and canine sleep macrostructure supports the idea that sleep reflects the dog’s wider context, not one isolated wellness variable.

Shelter research points in the same direction. In one small observational study, more daytime resting was associated with positive judgment bias, less repetitive behavior, and more staff-coded relaxed time, but the relationship between sleep and welfare was complex rather than linear. The shelter-dog resting and welfare study by Owczarczak-Garstecka and Burman does not show that a bed or a forced increase in rest causes better welfare.

A separate four-week telemetry study of eight male shelter dogs reported that social enrichment, especially female-conspecific presence, improved selected sleep-quality and day/night-cyclicity metrics, with substantial individual variation and no broad validation of the metrics as a welfare standard. The shelter telemetry study of enrichment and sleep indicators is best read as promising technical evidence from one small setting, not as a household prescription.

A practical bedtime routine for better dog sleep

The most defensible way to improve dog sleep is to make rest easier, more predictable, and physically comfortable while watching for medical causes of disruption.

1. Track the pattern before changing everything

   For seven days, record bedtime, wake time, major naps, nighttime awakenings, visible twitching, panting, scratching, coughing, accidents, unusual vocalization, and next-day behavior. Note exercise, training, meals, visitors, medication, temperature, and noise.

   This record helps separate a one-off event from a repeatable pattern. It also gives your veterinarian useful information if the problem continues.

2. Keep the final hour calm and familiar

   Use a predictable sequence: bathroom break, quiet movement, water access, a brief low-arousal interaction, and access to the usual sleeping area. Avoid turning bedtime into the most exciting part of the day. A calm routine does not need to be elaborate; consistency matters more than a special product or ritual.

   Research on pre-sleep activity and social experience suggests that what happens before sleep can influence subsequent sleep measurements. That does not prove that one exact routine will make every dog calmer, but it supports a measured approach: change one or two variables, then observe.

3. Improve the sleep surface and location

   The sleep area should be large enough for the dog to stretch, turn, and change position. It should have stable support, reasonable temperature, low traffic, and a surface that can be cleaned.

   

   For dogs that curl, lean, rest their chin, or prefer feeling surrounded, a bolstered bed such as the SnuggleSoft Orthopedic Calming Pet Bed may fit the physical preference described by its product design. It is a comfort option, not a treatment for pain, anxiety, or a sleep disorder.

   Dogs that naturally seek covered spaces may prefer a hooded cave-style pet bed. Let the dog adopt it voluntarily. Forcing a dog into an enclosed bed can create a negative association with the sleep area.

   A dog that leans, curls, or prefers bolstered support may also suit the ComfortCradle Orthopedic Dog Bed. Fit, placement, and the dog’s voluntary use matter more than the product label.

   For a room-by-room checklist covering bed size, placement, lighting, temperature, cleaning, and recovery spaces, use the Pet Comfort and Sleep Guide.

4. Protect sleep from avoidable interruptions

   Ask household members not to wake the dog for play, photos, or repeated checking. Keep the sleeping area away from doors, televisions, loud appliances, and high-traffic walkways when possible. If the dog sleeps in a crate, make sure the crate is already a comfortable, voluntarily used space rather than a new restriction introduced during a sleep problem.

   Do not wake a dog by touching them during a vigorous dream-like episode. Speak softly from a safe distance and remove nearby hazards. A startled dog may bite even if they are normally gentle.

5. Review comfort and health

   Pain, itching, respiratory disease, digestive upset, urinary problems, medication changes, and age-related cognitive changes can all affect sleep. A dog who suddenly changes sleeping location, struggles to lie down, repeatedly gets up, pants at rest, or becomes restless at night needs more than a new bed.

   The goal is not to force a higher REM percentage. It is to support normal rest while identifying anything that prevents it.

Can you measure REM stability at home?

You can monitor sleep patterns at home, but you cannot measure REM stability accurately without veterinary-grade physiological recording. A phone video, wearable activity tracker, or owner-created score may help document change, yet none should be treated as a validated CRSI or BSQR.

A useful home sleep log includes:

• Approximate time the dog settled.
• Approximate time the dog woke.
• Number of obvious awakenings.
• Twitching or vocalization: brief, repeated, forceful, or injurious.
• Breathing: quiet, labored, coughing, or interrupted.
• Morning behavior: normal, unusually tired, disoriented, irritable, or unable to perform familiar routines.
• Events that may explain the change, such as pain, heat, travel, visitors, exercise, medication, or noise.

Look for within-dog change rather than comparison with internet videos or another breed. A Labrador, Greyhound, French Bulldog, puppy, and senior mixed-breed dog may have different sleep patterns, and the evidence does not support a universal cross-breed REM ranking.

The same principle applies to behavior. A dog who sleeps less during one busy night is not automatically sleep deprived. A dog who repeatedly wakes, cannot settle, and shows a persistent daytime change deserves assessment.

When does dog REM behavior need veterinary attention?

Brief twitching is often compatible with normal sleep, but forceful, escalating, or seizure-like episodes should be recorded and discussed with a veterinarian. Seek prompt care if an episode involves injury, prolonged stiffening, repeated collapse, labored breathing, blue or pale gums, severe confusion after waking, or new neurological signs while awake.

A published case report described a 9-month-old Labrador retriever mix whose unusual sleep-associated episodes were confirmed by EEG to occur during REM sleep. The dog also had anxiety-related behavioral disorders. This case shows why EEG and clinical evaluation can distinguish an abnormal REM-associated disorder from seizure-like events; it does not show that ordinary puppy twitching is a disorder. The EEG-confirmed canine REM sleep disorder case report is a clinical example, not a prevalence estimate.

Before the appointment, take a video from a safe distance. Record the dog’s name, the date, how long the episode lasted, whether the dog responded to sound, what happened immediately afterward, and whether similar movements occur while awake. Do not restrain the dog or place your hands near the mouth during an episode.

Changes in behavior also warrant attention when they persist. Sleep may contribute to poor learning or emotional processing, but aggression, anxiety, irritability, or withdrawal can reflect pain, fear, neurological disease, training history, or environmental stress. Better sleep is supportive care; it is not a substitute for diagnosis or behavior treatment.

A better target than a perfect REM score

The most useful finding from canine sleep research is also the least sensational: dogs have measurable sleep stages, those stages show meaningful brain activity, and sleep interacts with learning, emotional processing, development, and cognitive health. The evidence is real, but it is more specific than many popular claims.

REM is one part of a larger sleep cycle. NREM contributes important learning-related activity. Age changes sleep architecture. Activity, social experience, location, and health can alter the pattern. A fixed CRSI benchmark across breeds does not yet exist, and BSQR is not a validated behavioral measure.

Protect your dog’s sleep by keeping the routine predictable, making the rest area comfortable, reducing avoidable interruptions, and tracking changes over several days. Use observations to guide a conversation with your veterinarian, not to replace one. Stable rest can support a dog’s ability to learn and cope with daily life, while persistent sleep or behavior changes deserve a full clinical view.