Assessing the Relationship Between Emotional States of Dogs and Their Human Handlers, Using Simultaneous Behavioral and Cardiac Measures

Abstract

Negative stress due to human handling has been reported for a number of domestic animals, including dogs. Many companion dogs display significant stress during routine care in the veterinary clinic, risking injury to staff and potentially compromising the quality of care that these dogs receive. On the other hand, positive interactions with humans can have a beneficial effect on dogs, particularly in stressful situations such as animal shelters. Research has shown that dogs can detect human emotions through visual, auditory, and chemical channels, and that dogs will exhibit emotional contagion, particularly with familiar humans. This study investigated relationships between emotional states of dogs and unfamiliar human handlers, using simultaneous measures of cardiac activity and behavior, during two sessions of three consecutive routine handling sets. Measures of cardiac activity included mean heart rate (HRmean), and two measures of heart rate variability (HRV): the root mean square of successive differences between normal heartbeats (RMSSD); and the high frequency absolute power component of HRV, log transformed (HFlog). We also assessed human handlers' emotional state during handling sessions following an intervention designed to reduce stress, compared with sessions conducted on a different day and following a control activity. Polar H10 cardiac sensors were used to simultaneously record cardiac activity for both canine and human participants, and behavioral data were collected via digital video. The strongest influence on the dogs' stress levels in our study was found to be increasing familiarity with the setting and the handler; HRmean and SI decreased, and HRV (as RMSSD) increased, significantly from the first to the third handling set. Canine HRV (as HFlog) was also highest in set 3, although the difference was not statistically significant. There were no strong patterns found in the human cardiac data across handling set, session, or by pre-handling activity. We did not find consistent support for emotional contagion between the dogs and their handlers in this study, perhaps due to the brief time that the dogs spent with the handlers. Recommendations for application to dog handling, and limitations of our methods, are described.