In patients with obstructive sleep apnea (OSA), nightmares, narcolepsy, parasomnias, and other sleep disorders, service animals may add a valuable, and currently underappreciated, treatment option.

By Mary W. Rose, PsyD, CBSM; Colleen G. Lance, MD; and Carlos H. Schenck, MD

As sleep specialists, we spend our patient contact time on issues related not only to adherence to the treatments we prescribe, but also on trying to understand the motivations and behavioral choices our patients make that affect their sleep. Part of this clinical endeavor is to further understand how we can use patient decision-making preferences to help guide and direct treatment, which at times can present a major challenge. In Western cultures, an increasingly prevalent patient choice is the practice of co-sleeping with pets.

It is the objective of this paper to describe not only the habits of co-sleeping, but also the use of service animals (SAs), usually dogs, and of emotional support animals (ESAs) as promising treatments of various sleep disorders, and to promote further interest in this area by sleep professionals. Not only are SAs apparently being more widely utilized for this purpose, but there may be other benefits to the treatment of sleep disorders that have been only initially explored. Dogs appear to be the most readily available and trainable SAs for most disorders. In recent years, there has been growing use of ESAs as well, who tend to be specifically used to provide support for psychiatric conditions such as panic anxiety and depression. Unlike SAs, these animals do not require formal training, nor are they legally protected for public access.

Historical Perspective

In order to properly understand the phenomenon of co-sleeping, we are best served by reviewing the history of our relationship with dogs and cats, as well as our (questionably unique) capacity of empathy for other species. Animal-human co-sleeping has a fascinating anthropological history. The earliest co-habitant of man was the dog. The domestication of the dog has been traced as far back as the Paleolithic era, about 31,000 years ago1 to 32,100 years ago.2 It is well-known that cultural patterns impact the sleep environment. In Western countries, the popularity of domestic pets has been increasing, and also rapidly expanding in westernizing cultures such as Japan. Over 70% of Americans own pets. Though most of the data on co-sleeping and bed-sharing lack large sampling or careful analysis, 60% of patients reported co-sleeping with their pets in one study.3 When a dog slept in the bedroom, it had a 57% chance of sleeping on the bed; 53% of co-sleepers reported sleep to be disrupted to some extent every night, but only 1% felt their sleep was disrupted more than 20 minutes per night. Yielding similar co-sleeping data, a recent abstract also suggested that 52% of patients sampled from a family medicine clinic indicated sharing a bed with a dog or cat; in spite of many patients reporting some disrupted sleep related to the co-sleeping, they also reported obtaining comfort and security.4

Sleep Professionals’ Perspective

Though the most common reaction of sleep professionals is to advise our patients against co-sleeping with pets, we fail to recognize that for many this is experienced the same as if they were told to stop sleeping with their spouse. Additionally, we automatically assume sleep disruption from co-sleeping with pets. However, this is not an evidence-based recommendation. Pets and/or SAs may truly be beneficial in the treatment of sleep disorders. Modalities discussed in this paper include: use of SAs to facilitate PAP use (and to diminish removal of the PAP at night); SAs to control nightmares in children; SAs to assist in narcolepsy therapy; and SAs to prevent wandering and injury in sleep walkers. Though dogs have certainly been used for years to assist patients with sleep disorders, most of this use was undocumented and did not involve formally trained SAs until the past few years. Currently, there are several trainers advertising the training of service dogs for this purpose, though there are no statistics on how many dogs are in use today for this purpose, how many trainers are available for this purpose, or clear guidelines on the training techniques and goals. Additionally, there is no universal requirement for who may train a dog or what the protocols should involve.

Obstructive Sleep Apnea

A case featured in a Houston Associated Press article in 1997 described a Chihuahua who would nudge or paw the patient when his PAP slipped off.5 There are several SA programs that include training of dogs for such purposes. However, as there is no registry of trained SAs, it is impossible to track how many have been trained for this purpose, what the training techniques are, or how effective they are. As discussed in the case study below, a SA was specifically trained to mitigate sleep walking, and to assist with PAP adherence, and has worked successfully for many years with our patient when all standard of care treatments had failed her. As most arousals that result in unconscious PAP removal appear to be brief, cuing from the dog with a paw touch seems to be just enough to successfully redirect the patient without waking him/her. Ideally, the SA would be trained to cue the patient, and in the case where the mask was removed, wake the patient to reposition the mask.


SAs may also be used to diminish nightmares and other anxiety and/or post-traumatic stress disorder (PTSD) sleep-related disturbances. Approximately 50% to 70% of PTSD patients suffer from chronic nightmares.6 Nightmares are complex; there is no single treatment that is effective; the causes are multifold and often complex, requiring individualized treatment protocols; and the patient must be more actively motivated and therapeutically engaged than with pharmacotherapy-based treatments alone. Nightmares have been associated with fear of sleep,7 which in turn perpetuates hyperarousal and insomnia. In the Veterans Affairs Medical Centers system, SAs are being more widely used to diminish the impact of PTSD-related nightmares through training in which they immediately awaken the nightmare patient and provide comfort, a role that no medication has yet to accomplish (unreported observations). Dogs are also used to mitigate anxiety, which is often associated with insomnia, and to modify hyperarousal and hypervigilance, which in turn creates a more amenable mood state for sleep initiation, as well as a greater sense of safety in those who are uneasy in the dark and/or night and who tend to phase-reverse to dodge nighttime sleep.

According to cognitive behavioral theories, the nightmare is perpetuated consequent to the maladaptive belief that the only way to escape the nightmare’s frightening consequences is through escape-avoidance. However, attempts to suppress unwanted thoughts prior to bedtime lead to intrusion of those thoughts into dreams.8 Desensitization through imagery rescripting and rehearsal is an effective strategy for diminishing the frequency and intensity of nightmares, suggesting that a proactive approach to managing these thoughts is highly effective.

Imagery rehearsal training (IRT) is a well-established technique for managing and helping overcome nightmares.9 It requires gradual exposure to the frightening nightmare material while being in a safe and nonthreatening environment (such as during the daytime and with daylight when the patient is relaxed and can review the content in a nonaroused state). This achieves desensitization as well as cognitive retraining. There is strong evidence from meta-analyses that image rehearsal in various forms (imagery rehearsal, exposure relaxation and rescripting therapy,10 and IRT with or without exposure to the original nightmare) has substantial benefits not only for reducing nightmare frequency, but with overall sleep quality, as well as for diminishing general PTSD symptoms (though the latter is not specifically targeted).9

The concept of diminishing escape-avoidance in order to maximize mastery over nightmares presents a conflict with the practice of training SAs to waken patients immediately when they appear to initiate the physiological symptoms of a nightmare. IRT encourages the patient to enter some level of discomfort into the nightmare imagery in order to gain mastery. If this is interrupted, it may perpetuate the escape-avoidance response and prevent mastery.

A combined approach of IRT with a SA might involve training the patient in the techniques of imagery rehearsal and training the dog to arouse the patient only when the patient reaches a heightened level or distressing arousal. This would allow achieving some level of mastery without perpetuating an escalated hyperarousal reaction. Additionally, it may be that for some patients the presence of the SA and the knowledge that they can more “safely” engage in IRT would make them more amenable to this treatment, and accomplish one of the necessary conditions of IRT, which is to be in a safe, reassuring environment.

There are multiple ways in which a SA may be effective in treating hyperarousal-related sleep disorders, with the dog providing an ongoing source of de-escalation of hyperarousal states. The dog thus serves as a constant and immediate cuing system. As noted earlier, dogs are also immediate warning signs that have no systemic negative effects (as might occur with medications).

Use of Family Pets for Nightmare Control

Author Rose finds that family pets, such as feline “Snicket” shown here, can diminish nightmares in children. Photo published with permission of child’s parent.

In the practice of one of the authors (MWR), family pets are often used for multifold reasons to diminish nightmares in children. The protocol involves the child sleeping with the chosen dog or cat and forming a narrative that involves nightmare rehearsal with rescripting to incorporate the pet into the modified nightmare scenario. Often the child will give the pet a designated magical power such as to fly with or to create a force field around itself and the child. As opposed to a more dramatic alteration of the nightmare, we opt for a progressive step-wise modification of imagery, with the exception of the pet’s power.

Children readily accept this concept and seem to effortlessly offer the power that they need their pet to possess. This provides the child with a conceptual escape route during a nightmare. The child’s nightmare in many ways mirrors the function of the fairy tale in highlighting developmental conflicts. This may be one reason why they can so readily identify the powers they would need to identify and overcome the adversaries in their nightmares. Their capacity for empathy and affiliation with their pets also allows children to easily identify with the pet and to imagine the powers they might attribute to the pet to assist them to modify and take control over the nightmare narrative.

Additionally, during an awakening, both the child and the pet remain together in the room and the pet can continue providing a sense of safety and protection against any lingering wake/sleep transitional monsters/threats that might have been perceived to come back into the waking world with them.


There are at least two agencies, Service Dog Academy in West Seattle, Wash, and PAWS’ Training Centers in four locations, that have trained numerous narcolepsy service dogs. Service Dog Academy reports having trained dogs to provide up to a 5-minute warning of an impending sleep attack so the patient can take precautions to minimize risk of injury during a fall. Dogs are also taught to place themselves in front of the patient to reduce the risk of falling onto a hard surface, to dial 911 from a specially designed phone, to fetch medication, and to cue patients to take medications to mitigate their disorder.11

The social and emotional effects of narcolepsy are known to be provocative for patients, and certainly an ESA may be as useful to these patients emotionally as are SAs for any chronic health condition.


Several case examples have been provided in which the efficacy of pharmacotherapy or hypnotherapy of aggressive and injurious parasomnias—REM sleep behavior disorder, sleep walking (SW), sleep terrors—was not only confirmed by the patient and spouse, but also by their pets, who eventually were willing to co-sleep with their owners again, once the pets were convinced that the owners had reestablished calm sleep.12 These pets not only had previously had their sleep disrupted, but at times had been injured from the owners’ aggressive parasomnia behaviors.

Case Example: We worked with a 54-year-old woman with multiple sleep disorders including OSA, sleep walking, and severe restless legs syndrome. During the CPAP titration, she required a full face mask due to nasal obstruction, which proved difficult as she struggled with severe claustrophobia that was likely related to abuse and trauma during childhood. Major complaints included nightmares and SW-related injury, along with dangerous behavior such as leaving the home and becoming combative toward her roommates if they interfered with her nocturnal wandering.

Despite minimizing the triggers and being on clonazepam therapy, the patient’s sleep walking continued. She left the home multiple times at night and sustained a leg fracture from falling. When her roommates attempted to redirect her, she responded with aggression. She was diagnosed and treated for OSA with CPAP but continued to have multiple sleep complaints, including no alleviation of SW episodes, frequent unconscious removal of her PAP at night, difficulty adjusting to the PAP mask, and continued nightmares.

The patient eventually located a SA program that was willing to undertake the training, which included teaching her dog multiple tasks: putting a paw on the patient’s PAP mask if she attempted to remove it at night; gentle redirection of the patient if she did remove her mask and rose from bed; and blocking of the patient if she attempted to leave the room or house while sleep walking. Consequent to use of the SA, the patient saw multiple benefits that she did not acquire during her pharmacological and psychological treatments pre-SA. The patient’s dog became not only a treatment provider for her sleep disorders but was considered by her to be a friend and a protector.


In several populations with sleep disorders, SAs may add a valuable, and currently underappreciated, treatment option. With some subpopulations, such as veterans, the significant overlap and high prevalence of sleep disorders such as OSA, REM sleep behavior disorder, and nightmares with comorbid PTSD could benefit substantially from such programs. Unfortunately, as there is no central registry for SAs, it is unknown exactly how many have been trained for specific disorders or how many requests for specialized training, such as for sleep disorders, have been made. This should be a top priority. However, it is essential that informed and motivated sleep specialists provide some input to these programs to facilitate the target goals.

The greatest risk or cost of SAs and ESAs is nearly always noted by the handler to be the financial cost of caring for the animal. In most cases, the handler considers the dog to be not just a worker, but a pet as well. For many, as the dog is not just a medical tool, these costs are assumed as they are for pets. Accessibility is less frequently cited as problematic as businesses in the United States are increasingly adherent to laws regarding SAs.

One of the greatest benefits of SAs, particularly for medically complicated patients, is that they offer a nonpharmacological solution to treating a sleep disorder in a way that may augment or even replace other (currently standard care) treatments. With regard to comorbid PTSD, the dog also offers a skill that is nonreplicable pharmacologically, eg, the dog can cue the patient consistently 24 hours a day based on scent and/or subtle behavioral cues prior to the manifestation of distressing symptoms, as has been demonstrated with epileptic seizures, as well as with cataplexy.13 Though the value of these service animals may be difficult to prove in a controlled experimental study, subjective reports not only appear to be highly promising, but patients seem to report long-lasting benefit through many years of their dog therapy.


Mary W. Rose, PsyD, CBSM, is assistant professor at the Baylor College of Medicine in the Sleep Disorders & Research Center in the Department of Medicine, Pulmonary, Critical Care & Sleep Medicine. She is also a licensed clinical psychologist in the Sleep Disorders Center at the Michael E. DeBakey VA Medical Center in Houston. Colleen G. Lance, MD, is assistant professor of medicine at University Hospitals Case Medical Center in Cleveland. Carlos H. Schenck, MD, is a member of the Minnesota Regional Sleep Disorders Center, a staff psychiatrist at Hennepin County Medical Center, professor at the University of Minnesota Medical School in Minneapolis, and a member of Sleep Review’s editorial advisory board. The authors thank Operation Wolfhound, 3732 W. Whitewater School Rd, Elfrida, AZ 85610, for offering information on their organization that was helpful to this article.


1. Germonpréa M, Sablin M, Stevens R, et al. Fossil dogs and wolves from Palaeolithic sites in Belgium, the Ukraine and Russia: osteometry, ancient DNA and stable isotopes. Journal of Archaeological Science. 2009;36:473-490.
2. Thalmann O, Shapiro B, Cui P, et al. Complete mitochondrial genomes of ancient canids suggest a European origin of domestic dogs. Science. 2013;342:871-874.
3. Shepard JW. Pets and sleep. Sleep. 2002;25:A250.
4. Duthuluru S, Stevens D, Stevens S. Sleep quality due to co-sleeping with pets. Sleep. 2014;37:A189.
5. Special guard dog keeps close watch on sleep apnea sufferer. Houston Associated Press. 1997;3-8.
6. Spoormaker VI, Montgomery P. Disturbed sleep in post-traumatic stress disorder: secondary symptom or core feature? Sleep Med Rev. 2008;12:169-184.
7. Davis JL, Wright DC. Randomized clinical trial for treatment of chronic nightmares in trauma-exposed adults. J Trauma Stress. 2007;20:123-133.
8. Taylor F, Bryant RA. The tendency to suppress, inhibiting thoughts, and dream rebound. Behav Res Ther. 2007;45:163-168.
9. Casement MD, Swanson LM. A meta-analysis of imagery rehearsal for post-trauma nightmares: effects on nightmare frequency, sleep quality, and posttraumatic stress. Clin Psychol Rev. 2012;32:566-574.
10. Davis JL, Wright DC. Exposure, relaxation, and rescripting treatment for trauma-related nightmares. J Trauma Dissociation. 2006;7:5-18.
11. Roy S. Service dogs support narcolepsy patients, literally and figuratively. October 31, 2014. Sleep Review.
12. Schenck CH. Paradox Lost: Midnight in the Battleground of Sleep and Dreams. Minneapolis: Extreme-Nights, LLC; 2005.
13. Dominguez-Ortega L, Díaz-Gállego E, Pozo F, et al. Narcolepsy and odor: preliminary report. Semergen. 2013;39:348-53. e41-46 Epub 2013 (Jul 5). English, Spanish.