Debra M. Eldredge, DVM
A recent study in Applied Animal Behaviour Science looked at the nocturnal activity and cortisol levels of dogs in shelters as potential indicators of their ability to adapt and settle into a shelter environment. To do this, they compared activity levels and urinary cortisol/creatinine ratios plus body weight between 29 shelter dogs and 29 pet dogs living at home. I freely admit, my first thought (having worked for two years in a shelter myself), was how well-funded and well-staffed this shelter in The Netherlands was!
Background for the Study
Any dogs with health problems, at age extremes (under a year or over 13 years) or showing obvious behavior issues were excluded from the study. Twenty of the shelter dogs were owner-relinquished, while nine were strays. When possible, intake history included whether the dogs had ever been kenneled. Dogs in the shelter were checked on days one and two, then day 12. Seventeen dogs who were adopted were also checked after being in their new homes for at least 6 weeks.
An earlier study at Cornell showed that shelter dogs were not bothered by lights being left on at night and tended to rest for a little less than an hour, then get up, adjust their position, and lie back down. That meant for this study, researchers had a basic idea about some shelter dog behaviors at night.
For activity, the dogs wore an accelerometer attached to a comfortable harness. Urine was collected free-catch on a walk in the morning or off the kennel floor if necessary. Dogs were weighed on the three designated days. They also had visual behavioral assessments.
Results
Not surprisingly, the shelter dogs showed more activity and cortisol on the first two days in the shelter compared to at-home dogs and to their own results done later. Interestingly, dogs who had been kenneled previously were more active than dogs who had not. This made me wonder if dogs who had not been kenneled were curled up, hoping to hide, while owner turn-ins were pacing and hoping their people would return quickly.
After 12 days at the shelter, the dogs had started to settle in and were producing less cortisol. Nocturnal activity had dropped a bit as well. “This study indicates that behavior testing should be delayed, or perhaps repeated, because you don’t want to delay the dog’s adoption. Presumably a dog would be better behaved after a good night’s sleep,” comments Katherine A. Houpt, VMD, PhD, DACVB, James Law Professor Emeritus, Section of Behavior Medicine at Cornell University. Certainly, dogs with extreme behaviors are likely to be identifiable on admission, but dogs whose initial fear and anxiety overcomes their normal behavior will tend to mellow with time.
What about small dogs?
Small dogs had higher levels of both activity and cortisol than large dogs. Perhaps they took the separation from their families harder or perhaps they had more stress from being in an environment with lots of strange dogs, many of them bigger – all just guesses on my part. Lena DeTar, DVM, DACVPM, DABVP-SMP, assistant clinical professor, Maddie’s Shelter Medicine Program, Cornell University College of Veterinary Medicine, says, “I don’t really know what to make of the finding that dog size is inversely related to elevation in urine cortisol levels and increased activity. Could it be that the smallest dogs don’t feel comfortable in a large room? Or that the largest dogs feel very constricted so don’t move as much? Does moving a lot help these smaller dogs, or contribute to their higher stress? The authors don’t attempt to explain why, which again I think is pretty good since it is unlikely to be anything more than a good guess.”
When to Evaluate Behavior?
My own take on this study is that it provides some black-and-white data to support what seems intuitive. Yes, dogs will be stressed on intake into a shelter. As they become more comfortable with the shelter routine and environment, objective indicators of stress should decrease. This also suggests to me that a true evaluation of a shelter dog for behavior should not be done right away. Sometimes there is no option, but if possible, dogs should have time to settle before evaluation. Dr. DeTar ruminates that it would be inhumane to wait two weeks to evaluate a dog for behavior when 1) formal behavioral evaluations are not extremely reliable 2) behavior questionnaires are more reliable and instantaneous, and 3) this study’s aims were not to find the point at which behavior returned to “normal,” but what stress levels might be like at the end of the two-week required holding period in The Netherlands. However, the finding that urine cortisol is already slightly lower on day two might further support the argument that waiting for three to five days (as some shelters still do) is unnecessary.
“The results of the study are likely to not be completely applicable to many shelters in the U. S. for a few reasons,” says Dr. DeTar.
“First, the description in the paper of the way that dogs are housed at the shelter in The Netherlands is quite nice compared to many shelters in the US in terms of construction, size, and enrichment plan. This is very important, because the holding period required as described for stray and surrendered dogs is long–up to 2 weeks–which we in the U. S. consider to be on the verge of long-term sheltering. Because of this, the time frames evaluated [first 2 days, 12 days] for the Dutch shelter dogs aren’t as relevant if our stray and administrative holds tend to be shorter, obviously a good thing, so we don’t know if/how dogs habituate in five to seven days the way they do in 12. Furthermore, the study doesn’t talk about “solutions” to shelter stress at intake beyond “habituation” to explain reduction in urine cortisol levels or ‘habituation vs learned helplessness’ to explain the lowered activity/longer rest findings. Essentially, this study just reports on the levels, which is actually great because too often work like this is over-interpreted.”
I also think this study backs up some premises for people who have a new dog or puppy – be it from a breeder, rescue, or a shelter. There is an adage that a new dog needs three days to begin to adapt, three weeks to start to feel comfortable, and at least three months before he is truly bonding to the family. The timing is different for an individual dog, but that is a decent guideline, and this study certainly supports the basic premise: “Don’t expect your dog to settle in immediately.”
As always, having a dog who is tolerant of being crated or kenneled is a plus. That dog will adapt faster and be less stressed if he is left at a shelter, boarded at a kennel, or needs to stay overnight at a veterinary hospital. Encouraging clients to crate-train their dogs is helpful for everyone.
Overall, Dr. DeTar felt this study had good attention to detail, with well-matched, well-designed, appropriate statistics. She would like to see the research continue with further tests that look at how dogs adapt and cope in different sheltering environments. For example, it would be quite interesting to look at different types of kennels and how single, double, room vs run-type enclosures contribute to rest for newly admitted dogs. This finding would certainly influence her consultation advice to shelters.
This article was reviewed/edited by board-certified veterinary behaviorist Dr. Kenneth Martin and/or veterinary technician specialist in behavior Debbie Martin, LVT.
