Exercise addicts need to stop talking about getting their endorphin hit, writes Hayley Bennett, and start looking at endocannabinoids

In 1979, health researcher Gary Egger was the sole subject in an experiment probing the connection between exercise and pain. He ran 20 laps of the University of Sydney’s cricket ground, measuring his pain threshold immediately before and after by strapping a blood pressure cuff around his arm and inflating it until it hurt so much he had to take it off. A glutton for punishment, Egger repeated this routine 15 times over six months, writing up the results with pharmacology colleagues for a short journal article: ‘The painlessness of the long distance runner’.

After a run, Egger could go six minutes before ripping off the cuff, compared to just four before running. As P L Pearson noted in the letters section, the pain-relieving effects were ‘fascinating’ and brought to mind anecdotal evidence of ‘ecstatic states… and anaesthesia’ associated with running. Later, these effects, along with sedation and reductions in stress and anxiety, became popularly known as ‘runner’s high’, though the term is too imprecise for some experts. ‘I don’t know what the evidence is for a “runner’s high”,’ says Cecilia Hillard, a neuroscientist at the Medical College of Wisconsin, US. She does, however, acknowledge the measurable effects of pain relief and elevated mood.

Despite scepticism about the term itself, scientists became absorbed in understanding what caused these effects, focusing on the possible contributions of endorphins – opiate hormones that hit the same receptors as painkillers like morphine. ‘Getting your endorphin hit’ entered the lexicon of sports enthusiasts. What hasn’t filtered through to the mainstream, though, is the fact that theories around endorphins have fallen out of favour in recent years. Meanwhile, new theories have materialised around another group of natural chemicals: lipid molecules called endocannabinoids that have an important influence on our brains and in which Hillard is an expert. Much as endorphins are considered as our internal opiates, endocannabinoids can be thought of as the body’s natural cannabis, with their receptors including the same one associated with the psychotropic effects of tetrahydrocannabinol (THC) from the marijuana plant.

Hillard’s career has spanned the entire history of endocannabinoids research, from their discovery along the path to understanding how THC works, to now. She was hooked from her first job as a research technician in 1977, which happened to be in the lab of pharmacologist Alan Bloom, whose team was working on THC. ‘That was way before we knew there was a receptor or anything,’ she recalls. ‘My fascination just got more and more intense as we figured out there was this whole signalling system that was huge. I mean, it’s everywhere and involved in so many different things.’

In the late 1980s, cannabinoid receptors were discovered in rats’ brains. We now know that cannabinoid receptor type 1 (CB1) – a G-protein coupled receptor – decorates the ends of neurons, where it can be activated by endocannabinoids (or THC) to mediate the release of the neurotransmitters that carry chemical signals between brain cells. The endocannabinoids themselves have turned out to be multifaceted molecules, playing a crucial role in brain development, and getting involved in everything from pain relief and stress management to hunger. Often, the endocannabinoid signalling system assumes a homeostatic role, ‘pulling us back to normal’, Hillard says. For example, it dampens the responses of other systems that produce the stress hormone cortisol. In relation to exercise, one key function is regulating energy intake by sending signals that encourage us to eat following a workout.

New highs

What specifically, though, connects the endocannabinoid system to the experience of a ‘runner’s high’? The narrative on endocannabinoids and exercise is relatively new, emerging only in the early 2000s. Neuroscientist Arne Dietrich and co-workers measured increased levels of the endocannabinoid anandamide in the blood of college students before and after an hour of running or cycling. They reasoned that, unlike endorphins, endocannabinoids cross readily from the blood into the brain – due to their lipophilic nature – and so could perhaps better explain the psychological benefits of exercise. This argument remains widespread, though as Gregory Rueggsegger, an exercise scientist at the University of Wisconsin–River Falls in the US, points out, the reality is more that with endocannabinoids, a rush in the blood gives us a better idea that brain levels are also up – whereas for endorphins, we don’t know, because they can’t cross over. ‘Some endorphins are produced in the brain; others are produced in your adrenal gland,’ says Ruegsegger. ‘So when you measure them in [blood] you’re essentially measuring that adrenaline-related response, but that can be completely separate from what endorphins are being made and how they’re acting in the brain.’

Alpha-endorphin - a 16 amino acid sequence of Tyr-Gly-Gly-Phe-Met-Thr-Ser-Glu-Lys-Ser-Gln-Thr-Pro-Leu-Val-Thr

Endorphins such as alpha-endorphin are no longer believed to be responsible for that rush after exercise

Nevertheless, the idea of endorphins as exercise-induced painkillers persisted, despite decades of research showing the opposite. The rebuttals came as early as the cricket ground paper. On five occasions, between running his 20 laps and strapping the cuff around his arm, Egger got a shot of naloxone – an opioid receptor-blocking drug usually given to reverse the effects of an overdose. The drug didn’t change the pain-relieving effects of his workout, suggesting that endorphins, as opioids, weren’t at play.

We measured the endocannabinoids in her system, and they definitely did go up with exercise 

One of Hillard’s Wisconsin collaborators – exercise psychologist Kelli Koltyn – also found that the pain-relieving effects of exercise were unchanged by naloxone, concluding that there must be something else besides endorphins. That’s when she turned her attention to endocannabinoids, as Hillard recalls. ‘We met up and starting talking, and measured the endocannabinoids in her system, and they definitely did go up with exercise and correlated with analgesia.’ In 2014, they showed that just doing a hand-grip exercise for three minutes could double blood concentrations of one endocannabinoid, 2-arachidonoylglycerol (2-AG), whilst also pushing up levels of anandamide and related lipids. Hillard suggests these endocannabinoids were produced in muscles, noting that whether muscle-derived endocannabinoids then cross into the brain to take their effects remains unknown.

A German group later concluded that the pain- and the anxiety-relieving effects of running require functioning endocannabinoid receptors – at least in mice, where they can be blocked with a drug similar to rimonabant, a CB1 antagonist previously marketed for treating obesity. However, rimonabant was banned over a decade ago due to side effects including anxiety, depression and suicide. That means there is nothing equivalent to naloxone available for blocking endocannabinoids in humans and it’s therefore hard to determine whether they are responsible for all the effects associated with exercise.

A tangled mess

So while we can consider the phrase ‘getting your endorphin hit’ a little outdated, it’s hard to simply replace endorphins with endocannabinoids. And there may be interplay between the two systems. Later work by Hillard and Koltyn showed that while blocking opioid receptors with naloxone doesn’t affect the overall benefits of doing the exercise, it does affect the anandamide (but not the 2-AG) response – instead of rising, anandamide levels stay the same. This suggested not only that opiate receptors and anandamide synthesis are somehow coupled, but that there is almost certainly more than one type of molecule producing the effects we experience from exercise. The endocannabinoid system must be working in concert with others, or the pain-relieving effects could be more easily obstructed. According to co-worker Kevin Crombie, now at the University of Alabama, the full complexity of the endocannabinoid system can’t be understood merely from measuring blood levels of endocannabinoids. ‘It’s nice that we do get this peripheral signal that is telling us something, but it’s not the end of the story,’ he says. ‘And it’s almost certainly going to involve other systems as well.’

The structure of 2-arachidonoylglycerol, an endocannabinoid which is an ester formed from the omega-6 fatty acid arachidonic acid and glycerol

2-Arachidonoylglycerol (or 2-AG) is one the endocannabinoids that could be responsible for ‘runner’s high’

Just to be clear, no one is arguing that endorphins don’t relieve pain: they do. It’s what they do in relation to exercise that’s at issue. Ruegsegger’s work focuses on the exercising brain and hints that the opioid system plays a different role to the pain-relieving one we’ve come to expect. Ruegsegger’s team looked at motivation levels in rats. They selectively bred female rats based on their activity levels and then compared the genetic profiles of ‘lazy’ versus active animals. Some of the biggest differences were in the opioid system, as well as the dopamine system – the brain’s reward system, which is tied to motivation for anything we might like doing, such as eating or exercising, as well as addictive behaviours like drug-taking. Manipulating either of these systems in rats produced similar effects. ‘If we blocked dopaminergic action, activity levels lessened, so dopamine is somehow tied to motivation for exercise,’ Ruegsegger explains. ‘Same story for opioids. So if you block that opioid receptor, you lessen the motivation for activity.’ However, when they blocked the rats’ opioid receptors after damaging their dopamine-releasing neurons, their activity levels didn’t change. The implication being that, in the brain, the opioid system is important for running – not so much for the ‘runner’s high’, but instead for activating the reward system tied to dopamine.

Virtually every single thing that we know causes our brain to say ‘do it again’ turns the dopamine circuit on, and endocannabinoids do too – but not as much as cocaine or nicotine

To make what Ruegsegger calls the ‘tangled mess’ of what happens during exercise doubly complicated, endocannabinoids seem to be plugged into this dopamine circuit too. Recent results even indicate the existence of a pathway linking cannabinoid receptors in the gut to brain-based dopamine signalling. According to Hillard, it’s a potential way that endocannabinoids in the periphery could influence the rewarding effects of exercise. While the connection is unconfirmed, there is already good evidence that endocannabinoids (and cannabis) can modulate dopamine-releasing neurons in the brain, though in milder way than many addictive drugs. As Hillard explains, ‘Virtually every single thing that we know causes our brain to say “do it again” turns [the dopamine] circuit on, and endocannabinoids do too, but they’re not nearly as big a wallop on that circuit as, say, cocaine or nicotine.’ Still, it goes some way to helping us understand why exercise can become addictive.

As the molecular details of endocannabinoids’ interactions remain somewhat fuzzy, blood endocannabinoid levels can still be helpful for giving us a system-wide snapshot of endocannabinoid activity in response to exercise. In 2021, the first comprehensive review on the effects of exercise on endocannabinoids, led by psychiatrist Hilary Marusak at Wayne State University in Detroit, US, brought together data from 33 studies. They found that it isn’t just running or resistance exercise that elevates blood levels. ‘The exercise didn’t seem to matter, so it could be swimming, resistance exercise… there are even a couple of yoga studies in there,’ says Marusak. ‘The take home message is “pick your poison” – whatever type of exercise you’d like to do, you’re probably going to get that benefit.’ They also found some evidence for a ‘sweetspot’ at moderate intensities compared to light or vigorous intensities. However, the impacts of long-term training were more difficult to fathom.

In the long run

A 2022 Spanish trial took a closer look at longer-term exercise effects on endocannabinoids in one of the largest groups so far – 102 people, who enrolled in a six-month long training programme involving aerobic and strength exercise. The trial was originally geared towards effects on fat stores, but ‘in parallel, we were really interested in understanding why exercise is beneficial – what is happening in the metabolism to make it healthy for you,’ according to researcher Borja Martinez-Tellez from the University of Granada. So they decided to measure levels of different lipids that might provide some clues, including endocannabinoids. Their results showed that people’s baseline endocannabinoid levels (measured outside of exercise sessions) decreased rather than increased during the training programme, but only with moderate not vigorous exercise.

If this seems counterintuitive, then Martinez-Tellez suggests thinking of exercise as a type of damage that induces inflammation and then repair. In this view, the body could adapt to the repeated release of inflammatory mediators – in which we can tentatively include endocannabinoids – by turning down baseline levels. Although the study seems to ask more questions than it answers about what endocannabinoids are doing, according to Martinez-Tellez, the fact that it’s moderate exercise that triggers both short and long-term effects could be telling us something about how to make the best use of a ‘runner’s high’.

However, as Hillard points out, endocannabinoid baselines can vary wildly from one person to the next, which makes it harder to predict the effects of changing them. ‘An analogy might be the level of water in a bathtub,’ she says. ‘If you have a bathtub that’s totally empty and you start putting water in, that’s going to have a different effect than if it’s nearly full and you start putting water in.’ Individual endocannabinoid baselines may have implications for our personal anxiety risk, which links back to the long-touted anxiety-relieving effects of running. The endocannabinoid system is thought to act as a ‘stress buffer’ and it’s interesting to note that the decline in endocannabinoids that happens during normal development – levels peak and drop off in adolescence – is linked to an increasing risk for anxiety. Concentrations of endocannabinoids and their receptors also appear to be altered in certain mental health conditions, indicating that life experiences can break the system.

There are so many unanswered questions

Meanwhile, endocannabinoid levels are also affected by our genes. We all produce an enzyme called fatty acid amide hydroxylase (FAAH), which degrades anandamide, 2-AG and other lipids. Some of us make a version that binds less well to its substrates, however, resulting in higher blood levels of endocannabinoids and, curiously, lower susceptibility to stress and anxiety. This has led to a flurry of interest in using FAAH inhibitors as drugs for manipulating the endocannabinoid system in stress-related disorders. ‘[The endocannabinoid system] is a really exciting system to target for conditions like post-traumatic stress disorder and also anxiety and depression,’ says Marusak, whose recent research looks at how natural variations in the FAAH enzyme affect anxiety risk in children. But she adds that because her work focuses on young people, drugs that could impact on brain development aren’t her first choice for manipulating the system. Exercise, though, may be a valuable tool. In an ongoing trial, Marusak is exploring how young people’s endocannabinoid levels and mood change in response to exercise, compared to stretching or meditation.

Exercise could also make a useful complement to existing PTSD therapies. Standard psychotherapies are based in part on the concept of ‘fear extinction learning’. They teach patients to down-train their fear responses when confronted with a harmless stimulus (often a picture) that in previous sessions was coupled to an uncomfortable electric shock. Early data suggests FAAH inhibitors can enhance fear extinction learning by increasing people’s anandamide baselines, with trials for PTSD now underway. But Hillard and Crombie’s team wondered whether exercise would have similar outcomes. In a 2021 study, they asked women with PTSD to walk or jog for 30 minutes after a therapy session, and tested their anandamide levels before and after the exercise. The results show that women who did the exercise at moderate intensity got a bigger boost to their extinction learning than those who did it at low intensity and, crucially, that their anandamide levels contributed to this boost. But, says Crombie, ‘There are so many unanswered questions, like: does exercise increase anandamide to the same magnitude as some of these FAAH inhibitors? It’s very early on and we are still working on elucidating the precise mechanisms at this point.’

The same could be said for ‘runner’s high’. We still don’t know exactly how it works. On the other hand, that doesn’t stop us appreciating its benefits. To return to Hillard’s original point: ‘I don’t know what the evidence is for a runner’s high. But I do know endocannabinoids go up and I do know they can elevate mood.’ And it’s these effects that we can now start to put to use, beyond the bounds of our daily workouts.

Hayley Bennett is a science writer based in Bristol, UK