The Effects of Therapy Dogs on Child Biology and Behavior

The central goal of the study is to determine whether brief interactions with a therapy dog have an immediate impact on children's biological response to stress, prosocial behaviors, and self-reported mood in comparison to interactions with a stuffed toy dog. The study uses a randomized crossover design with two study arms; all children will receive the both interventions during the same session, with the timing of the intervention randomized across subject. All outcomes will be assessed during a single study visit. No follow-up data will be collected. The study uses both between-group and within-subject comparisons. Between groups, the investigators predict that children who interact with a therapy dog prior to a psychosocial stress task (Arm 1) will show attenuated cortisol response to the stress task (primary outcome) and reduced physiological stress (secondary outcomes) compared to children who interact with a stuffed toy dog prior to the psychosocial stress task (Arm 2). As an additional control, children will also watch a 5 minute puppy video prior to the prosocial stress task and will then interact with a stuffed toy dog prior to the psychosocial stress task (Arm 3). Conversely, children who interact with the therapy dog immediately prior to the in-lab behavior tasks (Arm 2) will show higher levels of behavioral carefulness and prosocial behavior (primary outcomes) compared to children who interact with the stuffed toy dog prior to the behavior tasks (Arm 1). Within subjects across both study arms, increases in positive mood and decreases in negative mood (secondary outcomes) will be greatest following interaction with the therapy dog compared to the stuffed toy dog, after controlling for main effects of study arm. Within subjects, physiological markers of stress (secondary outcomes) will be lower during the interaction with the therapy dog than during interaction with the stuffed toy dog. Investigators will seek additional funds to collect and analyze salivary oxytocin data. The hypothesis is that children will show greater increase in oxytocin following interaction with the therapy dog in comparison to interaction with the stuffed toy dog. This study will also investigate the mechanisms through which child-dog interactions influence youth stress responsivity, using coded videotaped data from the subset of children in Arm 1 who interact with the therapy dog prior to the psychosocial stress task. It is hypothesized that child behaviors observed during the interaction, such as duration and frequency of eye gaze, petting and stroking behaviors, and use of positive affect, will be inversely correlated with change in cortisol response to stress. Dog behaviors, such as duration and frequency of eye gaze and approach behaviors, will be inversely correlated with children's change in cortisol response to stress. The study will also investigate whether child characteristics moderate the effects of the child-dog interaction. Investigators hypothesize that the effects of the therapy dog intervention will be stronger among children who currently live with dogs versus non-dog owning children and among children with more positive attitudes towards pets. It is also expected that the effects of the therapy dog interaction will be weaker among children with internalizing problems and for children experiencing higher levels of general stress. Investigators will also test whether the effects of the therapy dog intervention vary across child gender, race/ethnicity, or socioeconomic status, or child personality.