RESUMO
Four experiments were conducted to evaluate the accuracy and reliability of noninvasive evaluation of cortisol in saliva of dogs. In experiment 1, we measured the cortisol concentration in the filter paper on which 250-µL cortisol solutions had been quantitatively pipetted and in filter papers dipped in cortisol solution. In experiment 2, we collected the blood and saliva of dogs 3 times at 30-min intervals and compared the cortisol concentrations to examine whether the dynamics of cortisol in the blood and saliva are similar. The results of experiments 1 and 2 showed that the cortisol concentration can be quantitatively measured with this method and that the dynamics of cortisol concentration in the plasma and saliva collected by using filter paper are not different (P = 0.14 for experiment 1 and P = 0.51 for experiment 2). In experiment 3, to investigate the factors related to inducing stress in dogs by using the filter-paper method of collecting saliva, we compared the cortisol concentrations at 0 and 30 min after collecting the saliva of pet dogs. The dog owners completed a survey on their dogs, providing basic information and reporting the collection of their dog's saliva. We found that the cortisol concentrations increased significantly in dogs whose owners spent >2 min collecting saliva (P = 0.005), suggesting that prompt collection of saliva is necessary for accurate assessment of cortisol without induction of a stress response. In addition, the cortisol concentrations increased significantly in dogs whose teeth were not regularly brushed (P = 0.04), suggesting that regular teeth brushing mitigates the effect of the collection process on cortisol concentrations in the saliva, with minimal stress to the dogs. In experiment 4, we measured cortisol concentrations in pet dogs accustomed to having their teeth brushed by their owners, before and after interaction with their owners, to assess whether brushing induces stress in dogs. We detected that the cortisol concentrations significantly decreased after human-dog interaction (P = 0.008), suggesting that this method does not induce stress in dogs. Our study indicates that the method of saliva collection by using filter paper is effective in measuring the cortisol concentrations to evaluate stress, although certain steps are required to enhance accuracy.