Skip to main content

Verified by Psychology Today


Do Therapy Dogs Suffer from Stress When They Are Working?

Worries that therapy dogs get stressed during sessions are unfounded.

Howard County Library System Photo — Creative Commons License
Source: Howard County Library System Photo — Creative Commons License

It was during a conversation at a reception which included a number of psychologists and other social scientists when someone asked me about using therapy dogs to help people who were undergoing stressful situations. I described what was then an ongoing experiment that I was involved in. It was designed to test the effectiveness of therapy dogs when it came to relieving test-related stress in university students prior to an academic course examination. This study used a typical set up which involved a number of dogs and their handlers. They were gathered in an open area where university students who were anticipating a test in the near future could wander over and interact with the dogs. (Although the data had not yet been analyzed, it would ultimately show that such interactions did serve to reduce stress and anxiety in the participating students, and that the effects carried over for many hours.)

I had just finished my description of the study when one of the members of the group spun around to face me directly with an angry look. She was a sociologist and seemed to be enraged by the situation that I had described. At that moment she had one hand on her hip and was using the other to poke at my chest as she loudly announced "That sounds like the most unethical and inappropriate attempt at therapy that I can imagine. You take a dog into an unfamiliar setting, surround him with a bunch of strangers who are then going to rush up to him and grope him, touch him, stare into his face and babble at him. I don't know whether that kind of behavior is going to do anything to relieve the stress of the human beings who are participating, but it is certainly going to raise the stress level of any therapy dog who receives this kind of treatment. I don't know what kind of moral code you psychologists use, but it would seem to me that this kind of animal assisted intervention simply involves human beings benefiting from stress reduction at the cost of elevated anxiety and discomfort on the part of the therapy animal. I would certainly challenge the ethics of such forms of 'therapy'."

I tried to point out to her that the animals that were used as therapy dogs had all been pretested to make sure that they were comfortable with such interactions from strangers, and furthermore that the dogs and their handlers had all been trained and had received appropriate credentials. In addition, based on my own observations of my own dogs' behaviors during therapy sessions, it appeared to me that they seemed to like the attention they were receiving and they showed no visible anxiety-related or avoidance behaviors. She did not accept my arguments, claiming that I was biased because I was solely focused on the benefits that the humans got out of the situation and was completely insensitive to the distress suffered by the dogs.

Although it has now been many months since that confrontation I am happy to report that a body reliable data has finally entered the scientific literature which shows that therapy dogs do not suffer from any undue stress during therapeutic interactions with people. It was accepted for publication in the journal Applied Animal Behaviour Science. The research was done by a team of 14 researchers headed by Amy McCullough from American Humane with funding from the Human Animal Bond Research Institute (HABRI). It directly looked at the effects that participation in therapy sessions had on the emotional state of the therapy dogs. The study used therapy dog teams from five different hospitals scattered across the US. These hospitals were part of the "Canines and Childhood Cancer Study". This is one of the largest human-canine bond studies thus far, and it focuses on the impact that animal assisted therapeutic interactions have on children with cancer as well as the effects on their parents. This most recent study extends the measures taken to look at the impact that therapy sessions have on the participating dogs as well.

The research team gathered data in a number of ways. As a direct measure of behavior they videotaped each therapy session and coded the dogs' actions to look at affiliative (friendly and sociable) behaviors, and also coded any visible stress or anxiety-related behaviors on the part of the dogs. The frequency of each type of behavior was taken from the videos and a score was determined for each of the positive and negative responses. In total more than 400 videos of therapy sessions were coded.

These investigators also gathered information directly from the dogs' handlers including completion of the Canine Behavioral Assessment and Research Questionnaire (C-BARQ) when they first entered the study. This inventory looks at typical behaviors and reactions of each dog.

As a more targeted set of measures, after each therapy session the handlers were also asked to fill out behavior inventories regarding their dogs behaviors in the session as well as the nature of the activities which the handler and the dog engaged in.

Perhaps the most important measure that was taken during this study was the cortisol levels of each dog. Cortisol is a stress related hormone which can be measured by taking a saliva sample using a cotton swab. It is generally accepted that the higher the level of cortisol, the higher the level of stress that an individual is experiencing. In this study a total of nearly 600 saliva samples were collected and analyzed. For each dog a series of saliva samples was taken at prescribed times. The control or baseline samples were taken in the dog's home, one in the morning, another at noon, and another at night. There were two therapy related samples. One was when the dog arrived at the hospital and their therapy bandanna or service vest was put on. The researchers reasoned that if therapy sessions were usually associated with increased stress for the dogs, then seeing these signals that such a session was about to happen would trigger a rise in stress. Finally, approximately 20 to 30 minutes after the beginning of the animal assisted therapy session another sample was taken to see if the dogs were now experiencing elevated stress or anxiety levels.

The data are quite clear. Analysis of the video records showed few anxiety-related behaviors on the part of the dogs, but rather an awful lot of sociable and friendly activities. Most importantly there were no reliable differences in the salivary cortical levels of the dogs when comparisons were made between the baseline and the therapy-related test samples. In other words, the therapy dogs were not showing any increases in their stress levels related to their therapeutic activities.

I have no doubt that if you took a group of average, untrained, pet dogs into an unfamiliar environment and then crowded strangers around them who were trying to touch and interact with them, some of these dogs would certainly show stress-related and avoidance behaviors. But of course, the dogs used in therapy sessions are pre-selected and pre-trained. What this data clearly shows is that when they participate in a treatment session therapy dogs do not experience increased stress. Furthermore, the video records show that when they are providing therapeutic interactions the dogs show a continuous flow of friendly and social responses. In other words they seem to be enjoying themselves in the therapeutic situation.

Copyright SC Psychological Enterprises Ltd. May not be reprinted or reposted without permission


McCullough, Amy, (2017). Physiological and behavioral effects of animal-assisted interventions on therapy dogs in pediatric oncology settings. Applied Animal Behaviour Science,

More from Stanley Coren PhD., DSc, FRSC
More from Psychology Today
More from Stanley Coren PhD., DSc, FRSC
More from Psychology Today