In 2012, David Raichlen, an anthropologist then at the University of Arizona, put ten humans, eight dogs, and eight ferrets on treadmills. He drew blood before and after each animal ran, and measured levels of anandamide, a naturally occurring compound that activates the same brain receptors as cannabis. Running raised anandamide sharply in the humans and the dogs. In the ferrets, it didn’t move at all.
Why dogs and not ferrets
The distinction Raichlen was testing wasn’t about intelligence or domestication. Dogs, like humans, are cursorial animals, built by evolution for sustained travel over real distances. Ferrets are quick over short stretches but aren’t built to run far, and they don’t rely on distance running to survive the way persistence hunters or long-range foragers do. Raichlen’s hypothesis was that a neurobiological reward for running would only have evolved in species that actually benefited from doing it habitually, and the treadmill results, in a small but carefully controlled study, backed that up. The runner’s high, on this account, isn’t a universal mammalian response to exertion. It’s a reward system that seems to have evolved specifically in animals whose survival once depended on covering long distances on foot.
That framing connects to a broader, older argument in human evolutionary biology, sometimes called the endurance running hypothesis, which holds that humans developed a distinctive set of physical traits, long legs, short toes, an achilles tendon built for elastic energy storage, and an efficient sweating system, specifically to run long distances chasing prey that could outsprint us but couldn’t outlast us. If that’s roughly right, a body that rewards its owner chemically for doing the one thing it was substantially rebuilt to do starts to look less like a curiosity and more like exactly what you’d expect to find.
What’s actually happening isn’t quite what people assume
The popular explanation for a runner’s high has always been endorphins, but endorphins are large molecules that struggle to cross from the bloodstream into the brain, which made it hard to explain how something measured in blood could be driving a feeling in someone’s head. Anandamide is smaller and fat-soluble, crosses that barrier far more easily, and in Raichlen’s data, the size of the anandamide increase actually correlated with how much better people reported feeling afterwards, measured on a standard mood questionnaire. That correlation is the detail that turns this from a plausible story into an actual finding. The chemical most likely responsible for the runner’s high belongs to the same family of receptors that cannabis activates, not the opioid family people usually reach for when they explain the feeling to a friend.
It only shows up at a specific pace
A follow-up line of research from the same group narrowed things further. Walking for thirty minutes did nothing to anandamide levels. Running at maximum effort didn’t reliably produce the effect either. It was moderate, sustained jogging, roughly the pace most people would call a comfortable run rather than a sprint, that produced the biggest rise. That’s a narrower window than the phrase runner’s high usually implies, and it suggests the reward isn’t simply proportional to effort. There’s a particular intensity where it shows up most reliably, and both easier and much harder efforts miss it.
It’s worth being honest about the size of this evidence too. Raichlen’s original comparative study involved ten humans, eight dogs, and eight ferrets, a small sample by any standard, though the pattern has been picked up and discussed in enough follow-up work since 2012 that it hasn’t been treated as a fluke. Later reviewers have also noted that a second endocannabinoid, called 2-AG, doesn’t reliably rise with exercise the way anandamide does, which suggests the story is specifically about one particular chemical messenger rather than the whole endocannabinoid system responding uniformly.
What I’d take from this for an actual morning run
None of this guarantees a good run tomorrow. Individual variation here is real, some people report a strong, reliable lift from running and others rarely notice one at all, and the research doesn’t fully explain that gap yet. What I find genuinely useful in it isn’t a training tip. It’s the reframing. A morning run that leaves someone feeling clearer or steadier isn’t necessarily a triumph of willpower over a body that would rather still be in bed. In cursorial species, humans included, it may be closer to a body doing exactly what evolution built it to reward, at exactly the pace evolution built it to reward it at, on a schedule no one has to negotiate with willpower at all.