Salivary Cortisol Levels of Working Therapy Dogs

Document Type


Publication Date



Research documenting the level of work-related stress among therapy dogs is limited. This research was designed to measure salivary cortisol in working therapy dogs. Certified handler/dog (Canis lupus familiaris) teams were recruited to participate from teams currently volunteering in the Tri-Cities, Tennessee area. Male and female dogs of various sizes and breeds were recruited. Ten dogs contributed fivesaliva samples. Dogs were fasted for one hour prior to procedure. Samples were collected over a continuum of events, for comparison of salivary cortisol levels in multiple environments. Samples were collected as follows: (1) sample in the dogs’ homes, (1)sample upon arrival at the “work” place, (1) sample just prior to leaving the “work” place, (1) sample just prior to being groomed and (1) sample taken at the dogs’ regular veterinarians’ offices. This design provided 1 control sample, 2 work samples, and2 samples that may be indicative of elevated stress. It was hypothesized that: 1) Salivary cortisol concentrations will be higher in samples collected during grooming and at a veterinary clinic than in samples collected at home, and 2) Salivary cortisol concentrations will be lower in samples collected during “working” conditions than in samples collected during grooming and at a veterinary clinic. Saliva was collected by placing a Salimetrics Children’s Swab (P/N 5001.06) [dimensions 8 x 125 mm] into the dog’s mouth until saturated, or less than four minutes. After examination of the descriptive statistics of (n=10) across five different environmental conditions, it was decided that one of the cases represented an outlier and was removed from data as subsequent analysis revealed a cortisol level that was more than 23 standard deviations away from the mean. Three dogs had at least one sample with insufficient quantity of saliva for analysis, and were removed from the data. A one-way repeated measures ANOVA was conducted to compare salivary cortisol levels during different environmental conditions (pre-therapy, post-therapy, home, veterinary clinic, and groomer). There was not a significant main effect for condition, Wilks’ Lambda =.299, F (9, 2) =1.17, p=.51,multivariate partial eta squared = .70. The hypotheses were not supported. This research suggests that salivary cortisol of working therapy dogs is not significantly different than home, veterinary, or grooming conditions.


Johnson City, TN

This document is currently not available here.