Conditioned placebo effect in dogs decreases separation related behaviours

In humans, placebo effect can be produced by giving verbal information and also by conditioning when, after repeated administration of an active substance, an inactive compound that just looks like the drug administered before, can produce the effect of the active substance. Conditioned placebo effect has been reported in rodents, however, the dog (Canis familiaris) may also provide a promising model species. In our study dogs' behaviour was observed while they were repeatedly separated from their owners in the same unfamiliar room. First, subjects did not receive any pre-treatment (Baseline trial), then they participated in either of two different conditioning contexts: after having received either sedative drug (Conditioned group) or non-sedating vitamin (Control group) treatment, subjects participated in 3 conditioning trials on consecutive days. Finally, in the 'Test trial', both groups were separated from their owners after receiving placebo (non-sedating vitamin). Results show significant effect of the sedative drug conditioning; when comparing the change from Baseline to Test trials in the Conditioned and the Control groups, conditioned subjects showed less active signs of distress (U(26) = 48, p = 0.021) and more passive behaviours (U(26) = 50, p = 0.027). We also investigated the association between dogs' susceptibility to conditioned placebo effect and their expectancy bias towards positive outcomes and found a positive correlation (r(12) = 0.697, p = 0.008), suggesting that dogs with more positive expectations are more responsive to placebo treatment. Considering previous human findings about stronger responsiveness to placebo in optimistic people, our results support the validity of the application of a dog model towards a better understanding of some aspects of the placebo phenomena in humans.