Endorphins Get All the Credit. Is the Runner’s High Really Just about Them?

Ask what happens during exercise and the conversation soon reaches “endorphins.” The word can become shorthand for almost everything a workout is supposed to make us feel. What gets left out when one familiar chemical story stands in for the whole experience?

Exercise does not trigger one chemical switch. It produces a coordinated response involving signals that help mobilize fuel, regulate cardiovascular demand, shape pain and mood-related experiences, and communicate between tissues. Different molecules are measured in different places and on different timescales, which is why a list of “exercise chemicals” can be more misleading than helpful.^1,2,3,5,6

One early response involves the catecholamines epinephrine and norepinephrine, adrenaline and noradrenaline. Their circulating concentrations rise during many forms of exercise, generally increasing as the physiological demand becomes greater. These signals help support cardiovascular function and mobilize fuels, including glucose from liver glycogen and fatty acids from adipose tissue. During exercise, they are part of a normal physiological response to increased demand.¹

Other systems are more closely tied to how exercise may feel, but even here there is no single “happiness chemical.” Acute exercise can increase circulating endocannabinoids such as anandamide, although responses differ between molecules and studies. The endogenous opioid system also responds to exercise: in one PET study, high-intensity interval exercise produced changes in μ-opioid receptor availability consistent with endogenous opioid release, while moderate cycling did not produce the same overall receptor response. These findings help explain why the familiar endorphin story is incomplete, but they do not show that one molecule is responsible for a person’s mood or “runner’s high.”^2,3

Pain is another example of an outcome that cannot be reduced neatly to one signal. A 2024 systematic review and meta-analysis of 13 studies involving 498 healthy participants found that one session of aerobic exercise could temporarily raise experimental pressure-pain thresholds both near and away from the exercising muscles. Higher-intensity exercise produced the largest pooled effects, while low-intensity exercise had relatively little impact. The finding describes a temporary change in pain sensitivity; it does not mean that every hard workout acts like an analgesic drug or that one identified molecule explains the effect.⁴

Exercise also changes signals involved in communication and adaptation. Brain-derived neurotrophic factor, or BDNF, is involved in neuronal maintenance and plasticity, and a 2026 systematic review found a moderate, generally transient rise in circulating BDNF after acute cardiovascular exercise in the studies suitable for meta-analysis. Contracting skeletal muscle can also release signaling molecules commonly described as myokines. A 2024 review found that responses varied substantially across molecules and studies, with many changes appearing around exercise and then moving back toward baseline. Blood measurements can show that signaling is dynamic without telling us exactly what is happening in every tissue or which signal matters most for long-term adaptation.^5,6

That distinction between an acute signal and a long-term effect is important. A molecule may rise in blood without rising in the same way inside the brain. A receptor-binding change may be consistent with neurotransmitter release without proving that one chemical caused someone’s mood. And an acute increase after a workout is not automatically the mechanism behind the health adaptations that appear after months of training.^1,2,3,5,6

The body’s “internal pharmacy” is an appealing metaphor as long as it remains a metaphor. The next time a workout is reduced to a list of “happiness chemicals,” there is room to ask what was actually measured and how much of the story that measurement can tell.