August 31, 2017

Audience: Executive Leadership, Shelter/Rescue Staff & Volunteers, Veterinary Team

Organization: University of California
Investigator(s): Micaela Young
Grant Amount: $5,000
Project Type: Summer Scholar
Project Status: Research Complete

Project Summary

The Summer Scholar study aimed to determine if giving melatonin to shelter dogs in the evening would have an effect on their overnight activity, barking, or daytime behaviors. The study results show that melatonin at this dose had no clear effect on anxiety in the shelter environment. The groups differed significantly in two respects: the melatonin group was more active, and the melatonin group spent more time showing multiple defensive behaviors, the opposite of what was expected.


To determine if evening melatonin administration would have an effect on overnight activity, barking, or daytime behaviors in adoptable dogs in a municipal animal shelter


The study was a randomized double-blind placebo-controlled trial. Seventeen dogs from and open-admission shelter were randomly assigned into control and treatment groups with the latter being given melatonin at 0.1-0.33 mg/kg or identical milk powder-filled placebo capsules in food for four evenings in a row. Activity, barking and behavioral data were recorded. Subject enrollment and data collection took place over a two month period.


  • The melatonin treatment group was significantly more active than placebo group, and spent more time showing more than one defensive behavior.
  • No significant difference between two groups regarding:
    - barking, tail position and motion,
    - time spent displaying only 1 defensive behavior (e.g., growl, ears backward, tucked tail, etc.),
    - one or multiple displacement behaviors (e.g., looking away, licking objects, shake off, circle/pace, yawn, etc.),
    - or one or multiple social behaviors (e.g., approaching, sniffing, play bow, etc.).


Activity levels between the placebo-control and melatonin groups were significantly different, with the melatonin group being more active during the monitored time period, which ran from the evening melatonin administration to the morning when the shelter opened. This relationship was originally expected to be in the other direction with melatonin decreasing activity by increasing resting behavior. If it is a true association, increased activity could possibly be due to less behavioral inhibition in the environment due to decreased stress or perhaps improved quality of sleep the night before. Given there were some significant results, and the potential for confounding factors, this topic warrants further investigation into melatonin as a behavioral aide.