Monday, March 25, 2013

The Scientific Basis of Therapy Dog Work

Dr. Dawn A. Marcus of the Department of Anesthesiology of the University of Pittsburgh School of Medicine has written a review paper, The Science Behind Animal-Assisted Therapy, in the medical journal, Current Pain and Headache Reports, 17, 322, April 2013.  Dr. Marcus is the author of Therapy Dogs in Cancer Care, Springer, New York, 2012, and The Power of Wagging Tails: A Doctor's Guide to Dog Therapy and Healing, Demos Health, 2011. She is also the co-author of articles on use of animal-assisted therapy at an outpatient pain management clinic, use of AAT with fibromyalgia outpatients, and one about to be published on the use of AAT with migraine sufferers.  Citations can be found in the endnotes of her recent article.  (Sadly, Dr. Marcus has died.  It must be hoped that her wonderful research will be carried on by others.)

Dr. Marcus refers to studies “showing subjective pain reduction following animal-assisted interventions with therapy dogs,” and argues that “[p]hysiological changes have been identified in both humans and the dogs visiting them that support subjective impressions of reduced distress, decreased pain, and mood enhancement.”   Connecting self-assessments of the benefits of therapy dog visits with measurable physiological changes has long been a goal of those who believe a scientific foundation will reduce resistance from some quarters of the medical community to therapy dog work. Dr. Marcus acknowledges that “[b]ecause psychological and somatic symptoms may also influence the pain experience, these studies measuring subjective pain severity cannot determine if the therapy dog visit directly reduced pain perception or if pain impact may have been diminished secondarily because of other improvements.” 

Physiological Changes in Recipients of Therapy Dog Visits

Although evidence for therapy dog visits having “more than short-term entertainment value” is relatively thin, Dr. Marcus cites studies indicating that “therapy dog visits result in reductions in stress hormones, such as epinephrine and norepinephrine, as well as increases in endorphin levels.”  Oxytocin, which also provides anti-stress effects and increases pain threshold, has been shown “to increase following dog interactions.” Reductions in levels of serum cortisol and chromogranin A have also been demonstrated. Blood pressure has been shown to decline with dog visitations.  (For an earlier review paper concerning animal-assisted therapy, see Margo Halm (2008). The Healing Power of the Human-Animal Connection, American Journal of Critical Care, 17, 373.)

Physiological Changes in Therapy Dogs

Anyone who has handled a therapy dog for several years has been asked, as I have, if therapy work is good for the dog.  “It’s good for the patients, but isn’t it stressful for Chloe?”  Chloe and I have been a therapy dog team for nearly five years.  Indeed there can be stress for the dog. In a visit to a cerebral palsy facility, one child locked a hand onto Chloe’s back.  I kept her calm while two attendants gradually unlocked the vice grip of fingers. Fran Breitkopf encountered a similar situation in a memory care unit:

“Although Casey loved to visit people he was not able to sustain visits to deep dementia sections of Golden Hill. It exhausted him and he was only able to deal with the confusion that he felt for a short time.  I did have to pry a woman away from Casey in one of his visits to that section. She grabbed him and hugged him too hard. His eyes pleaded for help and we did have a nurse come in to help us. He was okay after that, not reluctant. When we left the facility, Casey would fall into a deep sleep by the time we got down the driveway to the main road, probably just 800 to 1,000 feet.  I ended up not taking him to that section after a time. On the other hand he loved going to Ten Broeck Elder Care Facility and was good for at least an hour. He just wanted to deliver kisses and sat on everyone's laps and gave them many. He did not want to sit on laps of people who had wet themselves. I always thought he didn't want anyone to think it was he who had peed.” 

Nevertheless, negative incidents are rare and Dr. Marcus describes a study in which visits to humans “resulted in significant positive changes (P≤0.01) in endorphin, oxytocin, prolactin, phenyl acetic acid, and dopamine levels in dogs.”  Corstisol levels have been shown to rise in dogs, however, indicating that there is a “need to limit visit frequency.”  Although my experience with Chloe has not been monitored for chemical changes in either of us, I do believe that the stress level varies with the assignment and that some assignments are actually fun for the dog. 

Mirror Neurons

In a section of her paper clearly inviting further research, Dr. Marcus considers the relevance of research on mirror neurons, i.e., neurons believed to be activated by watching activities that trigger responses similar to what is being watched.  This may explain why seeing disgust in someone else’s face may make a person feel a similar disgust, or watching someone vomit will induce queasiness. Dr. Marcus argues that such research may support “the hypothesis that humans witnessing ‘cheerful’ behavior in a friendly dog (such as attentiveness to the subject, mouth slightly open and ‘smiling,’ and tail wagging) might result in empathic imitation of cheerful behavior, possibly mediated by activation of mirror neurons.”

Therapy Dogs and Disease Detection

Dr. Marcus states that “anecdotal reports from therapy dog handlers suggest that, when therapy dogs are placed in a room with a number of people, the dogs tend to seek out for their attention those individuals who are ill or in distress.”  I have myself noticed this when there are both patients and visitors in a room, though I’m not sure that some level of cueing might be involved, i.e., that I want Chloe to focus her attention on the patient rather than the patient’s visitors, because that is why we are there.  If there were a group of people in a lobby and it was not apparent that one of them was a patient, I couldn’t describe her choices so easily.  Often it is the person that pays attention to her first that gets the attention from her.  Still, it is an interesting observation and deserving of more study. Needless to say, dogs get better at recognizing disease when they are trained to do so, as was demonstrated in one recent paper to which I contributed. 

Some research has indicated that stress responses in humans involve chemical changes, including “increases in glucocorticoids and catecholamines.”  Dr. Marcus notes that studies are not available to determine a dog’s response to such chemical changes, but that they may be able to detect them. 

Dog vs. Human Visits

Casey Visiting a Nursing Home on Halloween
Dr. Marcus summarizes research indicating that patients are more likely to look forward to receiving a dog visit than a visit from a friendly volunteer.  This conclusion will probably not surprise any therapy dog handler.  It is a joke among those handlers I know that neither the staff nor the long-term patients know us nearly as well as they know our dogs, and many do not even know our names or recognize us if we are not with our dogs.  Even patients with Alzheimer’s have been known to recognize a therapy dog by name.  Dennis Civiello, whose West Highland Terrier, Mackie, visits facilities in Phoenix and coastal Oregon, sent me the following account:

“In one case a patient that no longer recognized family members and didn’t move or speak would call out ‘Mackie’ when we arrived for our weekly visits. During the visit she made herself understood that she wanted Mackie on her lap and would attentively stroke and pet him. These moments demanded my focused attention and, more importantly, great care when taking Mackie from her lap because she resisted Mackie leaving. This experience remains vivid in my mind because of the patient’s reaction to Mackie and the reaction from family members that saw, in their eyes, life return to her.”

I once heard a social worker complain that there was no point in his visiting a memory ward when a therapy dog was there.  “No one wants to talk to you until after the dog leaves.” 


This paper presents a very sanguine perspective on the value of therapy dogs for hospital and other populations.  It is the perspective that I think will be established in the end.  I should point out, however, as I did in a blog last August, that there well-grounded reviews that have not been nearly as positive regarding benefits of various associations with animals.  More research is necessary.  I do not think that therapy dogs have to be justified by their providing long-term benefits.  Diana Edelman, in writing her poem to a therapy dog, once told me that she wrote the poem partially to fix Chloe in memory.  That was enough for her, and is enough for me and Chloe. 

Thanks to Fran Breitkopf and Dennis Civiello for contributing thoughts and recollections.   Fran's dog, Casey, a miniature poodle, was one of the best-known therapy dogs in upstate New York and died in January, having been a therapy dog for most of his more than 16 years of life. 


  1. Very interesting. Recent study on acupuncture suggests that the ability of acupuncture to reduce pain may also stem from lowering stress hormones.

  2. Thanks for the comments. By the way, love your book, which I always cite as the definite resource on legal aspects of service dogs, etc.

  3. What a well-written, informative, and touching piece! I've just begun looking into becoming a team with my newish rescue Westie, he is so energetic and licky. But we both need a lot of training. Your blog and this article have definitely given me incentive. When I think about the sadness and loneliness of illness, I remember my grandmother's final years and her greatest joy--sitting in her rocker with my poodle on her lap.

    I wish everyone in pain or loneliness could have their own therapy animal. Our shelters would be empty and our hearts would be full.

  4. Dogs must be trained to understand the demands of working with young children and the elderly.therapy dogs training