New study suggests touch is a key ingredient in dogs’ stress-buffering power

Stroking a pet dog may not immediately erase the physical toll of a stressful experience, but it appears to prepare the body to handle the next challenge more calmly. A new study published in Biological Psychology indicates that dog owners who physically petted their companions after a high-pressure mental test showed gentler brain and heart responses when facing a second round of stress shortly afterward. Simply sitting near the dog, making eye contact, and talking to it without physical contact did not provide the same protective effect.

Psychologists use the term stress buffering to describe how supportive social contact softens a person’s physical and emotional reaction to difficult situations. In human relationships, affectionate touch is a primary driver of this protection. In one frequently cited experiment, women who received a brief neck and shoulder massage from their partner before giving a speech showed smaller rises in heart rate and the stress hormone cortisol than women whose partners offered only verbal support.

Because many people treat their dogs as family members, scientists have examined whether companion animals offer similar biological protection. For instance, a study covered by PsyPost in August 2025 found that people accompanied by their dogs during a laboratory stress test had smaller spikes in the stress hormone cortisol and a more balanced nervous system reaction than those tested alone.

Isolating physical touch as the specific mechanism behind these benefits has been difficult, because owners typically pet, look at, and talk to their animals all at once. In a 2019 randomized controlled trial of university students, spending 10 minutes petting cats and dogs lowered salivary stress hormone levels more than sitting nearby and watching others pet the animals. However, trials testing whether animal touch calms people down during an active stressor have yielded mixed results. For example, a 2021 laboratory study of adolescents found that touching an unfamiliar therapy dog during a stressful evaluation did not reduce anxiety or physical arousal any more than talking to the dog without touching it.

Other evidence suggests that animal contact might instead act as a preventative shield before a challenge arrives. In a 2017 experiment with undergraduate students, interacting with a therapy dog beforehand muted the physical sweat-gland response when students later tackled a fast-paced math test. Building on this work, study authors Vanessa Kaufmann and Annett Schirmer of the Department of Psychology at the University of Innsbruck and Schirmer’s Brain and Behavior Lab, along with their colleague Xiaoqin Cheng, designed an experiment to isolate physical touch with one’s own dog from other forms of interaction, tracking both immediate stress recovery and resilience to a subsequent stressor.

Free daily newsletter

“Interactions with pets have a long history as therapeutic interventions and a lot of research suggests that they can indeed be good for us,” Kaufmann and Schirmer told PsyPost. “However, little work has been done to uncover potential factors or mechanisms that make human-animal interactions beneficial. Hence, we were interested in the ‘how,’ i.e., what is it about these interactions that might actually produce positive outcomes for the human?”

“Touch seemed like a particularly interesting candidate because physical contact emerges very naturally between people and their pets, and research on human touch has already shown that touch can buffer stress,” the researchers explained. “Therefore, we wanted to examine whether touch with a pet could have a similar effect.”

The researchers recruited 66 dog owners and their pets to visit a university laboratory in Austria. After excluding one person who stopped early and another who had excessive noise in their brain recordings, the final sample included 64 White European adults (51 women and 13 men) with an average age of 34 years. All of the dogs were healthy, over two years old, and comfortable in unfamiliar rooms.

Each participant completed two experimental blocks that followed the same five-step sequence: a five-minute resting baseline, a 10-minute stress test, a one-minute post-stress rest, a five-minute recovery period with their dog, and a five-minute post-recovery rest. During the 10-minute stress test, participants answered rapid-fire general knowledge questions within five seconds and counted backward in fixed steps to the beat of a flashing red dot. To induce social-evaluative stress, which is the anxiety of being negatively judged by others, participants were told their performance was being recorded for later analysis and were monitored over a video call by a staff member who gave only critical feedback.

The two blocks differed only during the five-minute recovery phase in a separate room. In the touch block, the dog roamed freely so the owner could spontaneously pet or hold it. In the no-touch block, the dog was kept on a short leash out of reach, preventing physical contact while still allowing the owner to talk to and make eye contact with their pet. Thirty-one participants completed the touch block first, while 33 completed the no-touch block first.

During the resting and stress phases, the researchers recorded brain activity using a cap with 64 sensors, a method called electroencephalography. They tracked alpha waves, which are slower electrical rhythms in the back of the head that accompany a relaxed mind and drop during stress, and beta waves, which are faster rhythms across the front and sides of the head tied to alertness and mental strain. Forearm sensors measured heart rate and heart rate variability, which tracks tiny beat-to-beat timing fluctuations driven by the calming parasympathetic branch of the nervous system, while self-report scales and video cameras captured participants’ mood and recovery behaviors.

Across both blocks, the mental test reliably triggered stress across multiple bodily systems. Compared to baseline rest, the stress task caused a drop in calming alpha brainwaves, a surge in alert beta brainwaves, a faster heart rate, and lower heart rate variability, with the rise in beta brainwaves lingering into the one-minute rest after the task ended. During the stress test, these physical responses synchronized with one another: stronger rises in beta brainwaves corresponded with faster heart rates, which in turn paired with lower heart rate variability.

By contrast, although participants reported feeling more negative, more tense, and less in control after the stress test, these conscious mood ratings showed almost no statistical connection to the size of their brain or heart changes. Only a small link emerged between higher heart rate variability and feeling calmer. The authors suggest this gap is consistent with research indicating that people’s reports of how stressed they feel may be tied more to context and expectations than to what is happening in their bodies.

When allowed to touch their dogs during the five-minute recovery break, owners spent about 64 percent of the time (roughly three minutes and 12 seconds) in physical contact, primarily stroking the animal’s fur. When the leash prevented touch, owners compensated by spending over 53 additional seconds making direct eye contact with their pets during the five-minute period compared to the touch condition.

“One notable aspect of the study is that participants were not instructed to touch their dogs,” Kaufmann and Schirmer noted. “When physical contact was possible, it emerged naturally, and participants spent about two thirds of the recovery period touching their pet. This highlights how central touch is to the way people naturally interact with their dogs.”

“Touch was very prominent in all the human-pet interactions we recorded, sometimes even when the pet wasn’t really into it,” the researchers added. “It seemed participants had a strong need to be in physical contact. Moreover, when this contact was not possible or allowed in the experiment, participants increased gazing at the dog, possibly as another way to connect.”

Contrary to the researchers’ initial prediction, touching the dog did not speed up immediate recovery: by the post-recovery rest, brainwaves, heart rates, and mood had returned toward baseline at similar rates whether the owner had petted the dog or not. The authors suggest the recovery setup may help explain why an immediate effect did not appear. Reaching out to stroke a moving dog involves physical activity that can mildly raise heart rate, and the extra eye contact in the no-touch condition may have offered its own calming effect by prompting the release of bonding hormones like oxytocin.

Instead, the benefits of touch emerged later, acting as a protective buffer when participants faced the second stress test. Participants who petted their dogs during the first recovery break showed a dampened biological reaction during the second stress test compared to their first stress test. Specifically, after recovering with touch, this group showed less suppression of calming alpha brainwaves throughout the second session, including during the stress test, along with a slower heart rate and higher heart rate variability during the second challenge itself, while also reporting a greater sense of control. Among participants who touched their dog first, the second-session differences had standardized effect sizes of about 0.36 for alpha, 0.40 for heart rate variability, and 0.72 for heart rate.

Participants prevented from touching their dogs during the first recovery break did not show these protective biological shifts during the second stress test; the differences in their alpha waves, heart rate, and heart rate variability between the two stress sessions were not statistically significant. Although this group reported feeling less tense after the second test, their beta brainwaves pointed to elevated tension during the second baseline period. Self-reported feelings showed an inconsistent pattern across the two groups and did not line up well with the physiological measures. Additional statistical models confirmed that the physiological results were not altered by the dog’s fur length, how much the dog whined or barked, or how worried owners felt about their pets.

“Our findings suggest that physical contact with a pet may reduce how strongly the body reacts to a subsequent stressful situation,” Kaufmann and Schirmer said. “Interestingly, there was no difference in how stressed participants reported feeling, suggesting that physiological buffering can occur without us necessarily being consciously aware of it.”

“While the effects were moderate in magnitude, we saw a consistent pattern across several physiological measures,” the researchers pointed out. “Brain activity, heart rate, and heart-rate variability all pointed toward a less pronounced physiological response to the subsequent stressor after participants had touched their dog. However, it remains unclear how effects observed in such a controlled laboratory setting translate to stressful situations in everyday life.”

The brainwave findings align with research covered by PsyPost in 2024, which similarly found that hands-on activities with dogs boosted relaxing alpha wave activity, though that earlier study measured immediate brain activity during calm interactions rather than buffering against a laboratory stressor. The authors speculate that stroking a dog’s coat stimulates specialized nerve fibers in human skin that respond to gentle, slow touch and send soothing signals to the brain. While these nerve fibers are scarce on the hairless palms of the hands, video recordings showed that owners mostly ran their fingers through their dogs’ fur, stimulating the sides and backs of the fingers where these receptors are present.

As with all research, there are some caveats to keep in mind. The study’s preregistered prediction, that touch would speed recovery right after the first stress test, was not supported. The protective effect on the second test came from an exploratory analysis, so it will need to be confirmed in a study designed to test it directly. In addition, the researchers cannot entirely rule out the possibility that touching the dog changed how much mental effort participants put into the second test, which could also lower physiological responses.

Participants also spoke out loud during the stress test but sat silently during rest. Because speaking changes breathing patterns, which directly influence heart rhythms, some cardiac changes during the stress test could reflect speaking or task effort rather than pure emotional stress, especially since a secondary, respiration-linked measure of heart rate variability did not replicate the main touch effect.

“Our findings provide causal evidence for a stress-buffering effect of touch under controlled laboratory conditions and as such support prior correlational observations that pet owners who engage in more physical contact with their pets report greater wellbeing,” Kaufmann and Schirmer told PsyPost. “However, real-life stress is varied and complex and it remains unclear whether different types of stress may differently benefit from touch. We also studied people interacting with their own dogs, so we cannot assume that the same effects would occur when interacting with an unfamiliar animal.”

The team also emphasized that the comfort of the animal is just as important as the comfort of the human. “Lastly, we would like to highlight the wellbeing of the animal who may or may not seek touch,” the researchers said. “Physical contact with another being should always be mutually wanted and positive.”

Sample makeup is another consideration, as around 80 percent of the participants were women and all were White Europeans. Future studies will need to test whether the findings hold across more diverse samples, examine other everyday stressors, and compare interacting with one’s own pet at home against interacting with an unfamiliar animal.

“A next step is to understand more precisely what it is about physical contact with an animal that produces these effects,” Kaufmann and Schirmer explained. “We are interested in both bottom-up mechanisms, such as how different characteristics of touch and the resulting sensory signals contribute to stress regulation, and top-down mechanisms, such as the relationship with the animal, expectations, or the meaning of the interaction.”

“One approach to tackle bottom-up mechanisms involves a closer look at how people touch,” the researchers continued. “Here we found that they primarily use their fingers relative to other body parts including the palm. Moreover, preliminary data suggests that contact with the fingers is particularly pleasurable, possibly because they are densely innervated by mechanoreceptors.”

The study, “Pet-induced stress buffering depends on touch: Evidence from brain and cardiac measures,” was authored by Vanessa Kaufmann, Xiaoqin Cheng, and Annett Schirmer.