Pet Ownership and Physical Health

Robert L. Matchock


Curr Opin Psychiatry. 2015;28(5):386-392. 

In This Article

Randomized Laboratory Studies

In laboratory studies, volunteer participants are typically randomly assigned to various conditions (some of which involve a dog or dog-like manipulation), either under resting or stressful conditions of varying degrees. Common dependent measures include cardiovascular indices such as BP and HR, hormonal measures (e.g., cortisol), self-reported stress and anxiety, and task performance.

Friedman et al.[50] first showed that exposure to a pet during a psychological challenge attenuates the stress response in a laboratory setting. Later, Allen et al.[51,52] found that participants exposed to a pet dog exhibited lower autonomic responses (i.e., HR, skin conductance, and BP) than participants exposed to a human friend during a stressful task. Since then, other researchers have also documented that social support in the form of a dog more effectively buffers one from stress than exposure to human friends.[53,54] Within the past several years, few laboratory studies have been published. Polheber and Matchock[54] compared three types of social support (human friend, pet dog, and a no support control) during the Trier Social Stress Test (TSST). They found that participants paired with an unfamiliar but friendly dog had attenuated HR and cortisol responses compared to participants paired with a close friend or participants with no social support. Lass-Hennemann et al.[55] exposed 80 women to a traumatic film clip either in the presence of a friendly dog, a friendly human, a toy animal, or while alone. Participants paired with either the friendly dog or the human had less self-reported anxiety and negative affect compared with the other two groups. However, components of the physiological stress response including HR, BP, and cortisol did not vary among the four groups. It is unclear why physiological parameters did not vary, but it is noteworthy that social support in the form of either a dog or a person effectively weakened the subjective stress response. It is tempting to conclude that the friendly dog and human conditions did not differ because there was no possibility for 'evaluation apprehension' in the current study as patients simply watched a traumatic film. In the Allen et al.[51] and Polheber and Matchock[54] studies, the stress manipulation involved a challenging task that also required performing in the presence of human friends. Similarly, Beetz et al.[56] tested 47 insecurely attached children with the child version of the Trier (TSST-C), randomly assigning children to an unfamiliar friendly dog condition, a toy dog, and a female graduate student condition. Salivary cortisol was significantly lower in the real dog condition compared with the other two conditions. Interestingly, the more time children spent petting the dog before the TSST-C, the lower the cortisol. This is important as it suggests a dose–response relationship between pet exposure and outcome measures.