Measuring Mechanisms of Stress Reduction in Animal-Assisted Interventions

• In Event: 1-047 - Animal-Assisted Interventions for Special Populations

Integrative Statement

Adolescence is a crucial period for the development of social anxiety at clinical and subclinical levels (Miers et al., 2013). One promising technique for reducing social anxiety is the use of animal-assisted interventions (AAIs). Emerging evidence suggests that AAIs may reduce anxiety, but three critical issues hamper the widespread use of AAIs as a clinical tool. First, little is known about which specific aspects of AAIs produce therapeutic benefits. Second, no studies have tested whether AAIs reduce anxiety specifically in adolescents with social anxiety. Finally, many studies rely only on self-reported anxiety measures using cross-sectional designs, limiting validity.

To overcome these barriers, the overall study aims to explore the specific mechanisms by which interacting with a therapy dog reduces anxiety, and test whether this interaction reduces anxiety in adolescents with varying levels of pre-existing social anxiety. Based on past research, we hypothesize an additive effect of social and physical interaction with a dog on anxiety reduction, such that participants with both social and physical interactions will show the greatest anxiety reduction, while participants with only the social interaction will show some benefit (compared to the control condition).

Using interim data from 34 participants (ages 13-17; 75 participants anticipated by 2019), this paper aims to demonstrate the feasibility and validity of 1) utilizing the Trier Social Stress Task for Children (TSST-C) to induce stress in adolescents across a range of social anxiety, and 2) using wearable physiology sensors (Empatica E4) to assess anxiety during an animal interaction.

Participants were randomly assigned to one of three conditions: 1) interaction with a toy dog (control), n=13; 2) social interaction (no physical contact) with a real therapy dog, n=10; and 3) social + physical interaction with a real therapy dog. n=11. Stratified random sampling was used to create a balanced distribution of individuals who range across low, mid, and high social anxiety (as defined by scores of >50 [high anxiety] and <36 [low anxiety] on the Social Anxiety Scale for Adolescents).

All participants completed the TSST-C, a laboratory-based social evaluative stressor with several phases: baseline, preparation/anticipation, speech and mental math (stressors), and recovery. We utilized a multivariate measurement approach with self-reported experience at each phase (State-Trait Anxiety Inventory) and continuous autonomic physiology (heart rate, electrodermal activity).

Results indicated that the TSST-C was effective in inducing significant changes in anxiety in adolescents across the range of social anxiety (Figure 1), as well as in each of the AAI/control conditions, (p<.001 for all groups). In addition, the E4 wearable physiology sensors were well-tolerated by the participants and also demonstrated patterns of physiological arousal during the stressor task (Figure 2). These results indicate construct validity, a necessary precursor to testing confirmatory hypotheses with the full sample.

This study represents a significant methodological innovation by using multiple measures (both self-report and physiological data) to test whether the social and/or physical features of AAIs reduce anxiety for adolescents. This work represents a critical step in translating a hypothesized effect into a new treatment option for those with social anxiety.

Authors

• Megan Kiely Mueller, Tufts University
  Presenting Author

• Heather Urry, Tufts University
  Non-Presenting Author

• Eric Anderson, Maine Medical Research Center Institute, Tufts University School of Medicine
  Non-Presenting Author