A Blog by

Through This Chemical Loop, Dogs Win Our Hearts

Wolves are wild, powerful, and fearsome predators, capable of bringing down even large prey. And yet, tens of thousands of years ago, some wolves started forming close associations with humans. They became more docile. Their bodies changed. They turned into domestic dogs. Today, we share bonds with them that can be as strong as those that tie us to other people. How did this happen? How did we go from fear to friendship? How did dogs start inspiring such genuine feelings of love and affection?

Miho Nagasawa at Azabu University, Japan has a possible answer. It involves oxytocin, a mammalian hormone that draws our attention to social cues. In yesterday’s post, I described mouse experiments which show how oxytocin sensitises an inexperienced mother to the distress calls of her pups, and eventually cements the bond between them. Dogs seem to have hijacked this chemical connection between mother and child, to cement a similarly strong bond with their owners.

Nagasawa showed that a dog’s gaze raises the oxytocin levels in its owner, prompting more social contact. In return, the owner’s gaze raises oxytocin levels in the pooch. This chemical loop unites the brains of two different species. “[It’s] a powerful mechanism through which dogs win our hearts—and we win theirs in return,” write Evan MacLean and Brian Hare from Duke University, in an accompanying editorial.

First, Nagasawa allowed thirty dog owners to interact with their animals for half an hour. She collected urine samples from both parties before and after that period, so she could measure the oxytocin levels in their bodies.

She found that the volunteers whose dogs gazed at them for the longest time experienced the biggest surges in oxytocin. They, in turn, spent more time looking back at their dogs, touching them, and talking to them. And the dogs that received the most reciprocal attention also experienced the biggest oxytocin spikes. Nagasawa had already shown the dog-to-human part of this loop in 2009 but she has now closed it, demonstrating that both species raise oxytocin levels in each other.

The same can’t be said for wolves. Nagasawa did the same experiment with eleven pure-bred wolves that were hand-reared by people. They weren’t pets, but they did have daily contact with their owners, who fed them and occasionally played with them. Despite their dog-like existence, these wolves did not make regular eye contact with their owners, and their gaze didn’t trigger a rise in oxytocin. The cross-species oxytocin loop only works between humans and domestic dogs.

Next, Nagasawa injected 27 dogs with oxytocin and placed them in a room with their owner and two strangers. After the injections, the dogs—but only the female ones—spent more time gazing at their owners, who then experienced a rise in oxytocin. It’s not clear why only the female dogs responded in this way. They might be more sensitive to the hormone, or less wary about the presence of unfamiliar people. Whatever the case, this second experiment confirmed that an oxytocin spike in one species can trigger a similar spike in the other.

These results offer important clues about the events that transformed wild wolves into domestic dogs. “During dog evolution, we have probably selected for a behaviour in dogs that elicits a physiological response in us that promotes bonding,” says Larry Young from Emory University. “That behaviour is eye-gazing.”

Among wolves, eye contact is a threat, which is why they rarely look directly at each other. Young suspects that wolf pups might communicate with their mothers through looks, triggering the same kind of affectionate cycle that exists in humans. “This just goes away as they mature,” he speculates. Perhaps as wolves evolved into dogs, they simply kept this child-like means of communication, just as they also retained some of the physical traits of their younger selves. In this way, they could have tapped into the oxytocin loop that strengthens bonds between human mothers and their babies, and triggered an almost parental affection.

Indeed, brain-scanning experiments have shown that there are overlaps in the brain regions that become active when human mothers look at images of their children or their dogs. “Diverse aspects of our biology appear to be tuned into dogs and children in remarkably similar ways,” write MacLean and Hare.

Reference: Nagasawa, Mitsui, En, Ohtani, Ohta, Sakuma, Onaka, Mogi & Kikusui. 2015. Oxytocin-gaze positive loop and the coevolution of human-dog bonds. Science http://dx.doi.org/10.1126/science.1261022

25 thoughts on “Through This Chemical Loop, Dogs Win Our Hearts

  1. My suspicion is that dogs adopted humans, not the other way around. Consider the legend of Romulus and Remus.

  2. Repeating the experiment with cats would be interesting. Even more interesting though would be to see if exotic animal (octopuses, snakes, etc) keepers have the same reaction to eye contact with their pets.

    Thinking about Chunfen Zhou’s question, would eye contact with pet pigs vs farm raised pigs show something similar. Basically determining, if the reaction more or less conditioned and/or somewhat voluntary.

  3. Follow confirmation links appears to be broken for Phenomena. Confirmation for this and one for No Place Like Home fail to link to the confirmation site.

  4. This seems like a chicken & egg conundrum – looking into your dog’s eyes helps cement the bond between the two of you, but that wouldn’t actually occur until you had already formed a bond, i.e. domesticated the animal. It’s a shame this study didn’t measure the effects of the “unknown” individuals on oxytocin levels – I wonder if the data was bad, it seems strange that they would have included them at all in the study but not collected relevant data.

  5. I have 3 Keeshonden and love them very much. I am glad to know that OT is involved. I understand that OT trials have been performed in autistic persons to try to raise their level of social interactions. It sounds promising to me.

  6. > I was wondering if the results could be replicated among people from a dog-eating culture…

    Well, why not? Some people (used to) have a close bond with dogs, and then once it dies, it becomes soup. (If you think that’s cruel, substitute an ox or a goat and see if it sounds OK now.)

    (Then of course there are people who just don’t like dogs, and people who only likes dogs if served in a bowl…)

    As an even more absurd example, there are a tribe where the relatives of the deceased perform(-ed?) ritualistic cannibalism (which led to Kuru disease). I’m pretty confident that they mourn their dead the same way all other people do.

  7. I wonder if the reaction is the same for all mammals that we have domesticated. I have formed amazing bonds with cats, dogs, rats and horses. And I do spend a lot of time looking into their eyes.

  8. If only eye contact is what that triggers oxytocin that bonds, what about blind people or blind dogs? We can see many loyal dogs whose owners are legally blind.

  9. Cool article, however, for the sake of science, let us remember that dogs are not descended from wolves. Dogs and wolves share a common, now-extinct ancestor, though to be a wolf-like canid. It’s easy to say “wolf” to describe this ancestor because they most likely physically resembled the extant Grey Wolf, but the common ancestor of today’s wolves and today’s dogs would have been genetically different from both species. Scientists do not believe that dogs descended from the wolves we know today, they believe that Canis Lupus, the Grey Wolf, and Canis Familiaris, the domestic dog, descended from a common, extinct ancestor.

  10. There would be no dogs without humans. Some wolves and some humans adopted each other. And it was a successful match. Evolution took care of the rest.

  11. does this mean that there is a genetic connection between being docile and having eyes that create this response?

  12. I have noticed that whenever I made eye contact with a dog, it often looks intently back at me. Even when it is walking with it’s owner it will walk on and keep looking back. I have found this curious. Also how if I am driving by and I catch the gaze of a strange dog, it will keep looking on and will respond by sometimes even stopping to look back.

  13. I don’t agree with the researcher’s theory of how and why dogs evolved to have this response to humans. I would like to see studies replicated with ‘domestic’ dogs that have different associations, training levels and experiences with humans, to be sure that the oxytocin isn’t just a type of ‘pavlov’s dog syndrome’ whereby the eyes are just a learnt signal/precursor to a positive interaction or reward. Just because humans respond a certain way to their babies and pets does not mean it is the same between dogs and their offspring. Wolves/Dogs/Canids rarely look at their young, and eye contact often is associated with warning or expectancy of a cue, so to assume that the oxytocin increases are due to the same emotions in both species is quite a stretch and not backed up by scientific evidence here. It is a shame if they didn’t think of this in their sampling which leads to the results of this single study being promoted as truth, as I still don’t believe it’s a good or positive thing to stare at a dog (based on their natural history) 🙁

  14. This is no surprise to me. I have always noticed the warm fuzzy feeling I get when I gaze into the beautiful eyes of one of “furry kids” .

  15. I’m Asperger. I adore my dogs and we live as “one being.” They aren’t perfectly behaved and neither am I. I don’t feel the same way about people. How does oxytocin explain that?

  16. Trainers advise that one should not stare into the eyes of a unknown dog as this will be interpreted as aggression. Perhaps the oxytocin effect occurs only with familiar dogs. As a volunteer with a group that rescues urban strays, I would advise caution about direct and prolonged eye contact with a strange dog.

  17. Interesting research, and interesting article … but it didn’t answer the main question I had: Was the research conducted well? Did the researchers publish their data for review? Did they state the theory they were trying to prove or disprove prior to conducting the research, or did they come up with a media-friendly theory after picking through the data? Our current research apparatus is biased towards eye-catching press coverage, exactly like this article. Peer review alone is not working. Perhaps it’s time for serious science journalists to interject some perspective into each and every single article that covers research. I understand that even good journalists like Ed Yong can’t replace peer review, but when the majority of peer-reviewed papers in legit journals are later found to be false, it’s appropriate to caution readers and give some context. Does research with this sample size generally hold up 1, 2 or 5 years later? Was data published? Was the theory stated prior to research? Etc. Thanks in part to Ed and others, the audience for science is ready for that kind of information. Thanks!

    1. @James Cage:

      Did you read anything beyond the headline? Almost all of your questions are answered in Ed’s summary, and the ones that aren’t are answered in the journal article that is linked at the bottom, of which you seem to be questioning the existence.

      Also, I’d like to know your source for “the majority of peer-reviewed papers in legit journals are later found to be false,” because as far as I know, retraction rates for peer-reviewed, legit journals are well under 1%. Although, to be fair, retraction rates for non-legit journals are 0%.

  18. RE: “Nagasawa injected 27 dogs with oxytocin and placed them in a room with their owner and two strangers. After the injections, the dogs—but only the female ones—spent more time gazing at their owners, who then experienced a rise in oxytocin. It’s not clear why only the female dogs responded in this way. ”
    This might explain it. “Oxytocin plays an important role in the neuroanatomy of intimacy, specifically in sexual reproduction of both sexes, in particular during and after childbirth. It is released in large amounts after distension of the cervix and uterus during labor, facilitating birth, maternal bonding, and, after stimulation of the nipples, lactation. Both childbirth and milk ejection result from positive feedback mechanisms.” – Marieb Human Anatomy & Physiology 9th edition, chapter:16, p. 599

Leave a Reply

Your email address will not be published. Required fields are marked *