The runner's high has been explained by endorphins for decades. Two studies have moved that explanation. The first put humans, dogs and ferrets on treadmills and measured what appeared in the blood. The second worked out which receptors the effect actually runs through. What they show together is narrower than the headlines about it, and more interesting.

Three species on the same treadmill

Dogs are running animals. Ferrets are not. Zoology calls them cursorial and non-cursorial mammals respectively: animals built for sustained locomotion and animals that are not. Humans belong to the first group and share a long list of morphological traits with it.

That difference is what Raichlen and colleagues built an experiment around, published in the Journal of Experimental Biology in April 2012 under the title "Wired to run". They measured circulating endocannabinoids in humans, dogs and ferrets before and after treadmill exercise.

The result has three parts, and the second is the one usually dropped:

1. Humans and dogs showed a significant increase in endocannabinoid signalling after high-intensity endurance running.

2. The same species showed no significant increase after low-intensity walking.

3. The ferrets showed no significant increase after exercise at any intensity at all.

So it is not movement as such that raises the levels. It takes endurance work, and it takes an animal built for it. The authors' own conclusion is that this is the first evidence that variation in neurotransmitter signalling between species may explain differences in how they move, and that a neurobiological reward for endurance exercise may be why humans and other cursorial mammals run habitually despite the energy cost and the injury risk.

The mouse that showed where it sits

A rise in the blood is not proof that the rise causes anything. That step was taken in 2015, by Fuss and colleagues in PNAS, under the blunter title "A runner's high depends on cannabinoid receptors in mice".

The group combined pharmacology, molecular genetics and behavioural experiments in mice, and broke the runner's high into components rather than treating it as one feeling. Three of them could be measured in the animal:

  • Reduced anxiety. Depends on intact CB1 receptors on forebrain GABAergic neurons.
  • Reduced pain. Depends on activation of peripheral CB1 and CB2 receptors.
  • Sedation. Was affected neither by cannabinoid receptor blockade nor by opioid receptor blockade.

The last line is worth pausing on. This is not a study that dismisses the opioids across the board. Running raises blood levels of both beta-endorphin and anandamide, which the authors state themselves. What the experiment shows is which receptor blockade removes which component, and for sedation neither removed anything.

The part no mouse can answer

Here comes the sentence that ought to appear in every summary of the study and almost never does. The authors write plainly that euphoria cannot be studied in mouse models.

Euphoria is at the same time the component most people mean when they say runner's high. The sudden feeling of elation is the whole reason the phrase exists.

So what has been shown is that two of the components, the anxiety reduction and the pain relief, depend on cannabinoid receptors in mice. The third, sedation, appears to run through neither cannabinoid nor opioid receptors. And the fourth, euphoria, is unanswered in both directions. The study neither confirms nor refutes the endorphin explanation for that part. It could not test it.

It is a distinction that disappears the moment the finding is retold, and it is the whole difference between knowing something and having swapped one story for another.

What the studies do not show

Raichlen's study is comparative and measures circulating levels. A level in the blood is not an experience, and the design cannot separate cause from effect in the species that responded.

Fuss's study is done in mice. The conclusions about individual receptor populations rest on blockade and knocked-out genes, that is, on interventions that cannot be performed in humans.

And neither says anything about introducing something from outside. Both are about the body's own molecules under the body's own load. The closest thing to a practical conclusion is that the usual explanation for why running feels good is less well founded than it sounds, and that the species with no need to run does not get paid for it either.

Read on

What chronic stress does to the endocannabinoid system, on the same system under the opposite load.

This article describes published research on the body's own endocannabinoid system. It is not medical advice and contains no claims about health effects of cannabidiol or any other product.