Abstract
The research into human-canine interactions (HCIs) has grown substantially, yet limited attention has focused on the welfare of canines involved, particularly pet dogs owned by volunteer participants. To address this gap, we conducted a secondary analysis of data from a randomized controlled trial, examining canine welfare during an acute human stress protocol. Our methodology incorporated evidence-based screening tools, environmental modifications, researchers trained in canine behavior assessments and safe interactions, and canine stress monitoring using the Fear Free™ Canine Fear, Anxiety, and Stress (FAS) Spectrum. Dogs' stress levels showed a non-significant increase from the rest to stressor phase (0.80 to 1.00, p = 0.073) and a significant decrease during recovery (1.00 to 0.48, p < 0.001). Only two dogs (7.6%) required withdrawal due to elevated stress levels, though these levels remained within acceptable safety parameters. The peak stress remained within acceptable limits, with only 24% (6 of 25) reaching an FAS score of two during the TSST. By final recovery, 96% of dogs achieved FAS scores of zero to one (Green Zone), indicating relaxed states. Salivary collection proved challenging, highlighting limitations in low-invasive physiological measurement techniques. Based on our findings and literature review, we propose standardized guidelines for HCI research, including thorough pre-screening, environmental preparation, researcher training, stress-monitoring protocols, and informed consent procedures emphasizing withdrawal rights. These guidelines aim to establish ethical standards for this rapidly expanding field, protecting canine participant welfare while enabling valuable research to continue.