Research into the effects of animal-assisted interventions (AAIs) has primarily addressed human health outcomes. In contrast, only few publications deal with the therapy dog experience of AAIs. This paper provides an overview on potential welfare threats that therapy dogs may encounter and presents the results of a review of available studies on welfare indicators for therapy dogs during AAIs. Previous investigations used physiological and behavioral welfare indicators and dog handler surveys to identify work-related stress. Research outcomes are discussed in the light of strengths and weaknesses of the methods used. Study results suggest that frequency and duration of AAI sessions, novelty of the environment, controllability, age and familiarity of recipients modulate animal welfare indicators. However, this review reveals that currently, clear conclusions on how the well-being of dogs is influenced by the performance in AAIs are lacking due to the heterogeneity of programs, recipient and session characteristics, small dog sample sizes and methodological limitations. This paper further aimed to identify unresolved difficulties in previous research to pave the way for future investigations supporting the applicability of scientific findings in practice.

Beneficial effects of human-animal contact on human health have contributed to the wide distribution of animal-assisted interventions (AAIs). While considerable effort has been devoted to the study of human welfare during AAIs, potential effects on therapy animals have been addressed less frequently. The aim of this study was to determine baseline and work-related levels of cortisol, a glucocor-ticoid hormone that mediates physiological responses to arousal, in certified therapy dogs and therapy dogs in training. All dogs (n = 21) participated in weekly group-AAIs in adult mental healthcare. Saliva samples were collected over the course of AAIs and on non-working days and analysed with a cortisol enzyme immunoassay. Analysis of the results revealed that according to their cortisol responses, both therapy dogs and therapy dogs in training were not stressed by AAIs. However, cortisol levels during work in certified therapy dogs performing AAIs on- and off-lead varied significantly, suggesting that further investigation into the use of a lead or other methods of giving therapy dogs opportunities to approach or avoid human contact is needed.

During the past decade, the field of human–animal interaction(s) research has been characterized by a significant increase in scientific findings. These data have contributed to our current understanding of how humans may benefit from contact with animals. However, the animal experience of these interactions is still an under-researched area. This paper addresses the welfare of dogs who participate in animal-assisted interventions (AAIs) to improve health in human recipients. This paper builds on previous work by Glenk (2017) and provides an updated review of the literature on therapy dog welfare published from 2017–2021. New advances in scientific methodology, such as the determination of salivary oxytocin, breath rate and tympanic membrane temperature, are analyzed regarding their value and limitations for research in AAIs. Moreover, welfare-related social and environmental factors (e.g., freedom of choice, exploration of novel environments, inequity aversion, individual development, working experience, relationship with handler and handler skills) that profoundly influence dog perception and well-being are reviewed and discussed. Accounting for the globally increasing interest and the number of dogs utilized in AAIs, safeguarding therapy dog well-being, and identifying situations, circumstances and protocols that may challenge animal welfare remains an emerging and crucial area of scientific effort.