November 2026

Five body chemicals get blamed for every mood. The evidence says less.

I tested the popular story that dopamine is pleasure, serotonin is happiness, oxytocin is love and cortisol is the enemy. Each chemical does real work. Each is also oversold, and most cannot be measured at home.

01 · The map: two systems, five names

Most of what gets called a “chemical reaction” in the body runs through two systems. The adrenal medulla releases epinephrine (adrenaline) and norepinephrine within seconds and drives the fight-or-flight response.1 The HPA axis, which runs from the hypothalamus to the pituitary to the adrenal cortex, releases cortisol over minutes to hours and reaches the heart, blood vessels and brain.2 Oxytocin, serotonin and dopamine are different cases. They work mainly as signals between nerve cells, and in the case of serotonin most of the body’s supply sits somewhere else entirely.

Five chemicals compared by source, main targets, speed and the popular label that misleads
ChemicalMade mainly inMain targetsSpeedPopular label that misleads
AdrenalineAdrenal medullaHeart rate and contractility, blood vessels, pupils, kidney renin release3Seconds“The fear hormone.” It also rises with exertion and stimulants.
CortisolAdrenal cortex, driven by the HPA axisBrain feedback loops, metabolism, immune function, blood vessels and heart2Minutes to hours, with a daily rhythm4“The bad stress hormone.” The body needs a morning rise.
OxytocinHypothalamusBrain circuits for social cognition and behavior5Varies with context“The love hormone.” Effects depend on context and person.
SerotoninAbout 95% in the gut, a separate pool in the brain6Gut, immune function and metabolism in the periphery; mood circuits in the brain7Varies“The happiness chemical you can eat.”
DopamineMidbrain neuronsStriatum and cortex, for motivation, learning and reward prediction8Sub-second bursts and slower tone“The pleasure chemical.”

Snapshot of the research as of November 2026. The science on each of these keeps moving, and several findings below rest on small studies.

02 · Adrenaline: the fastest signal

Epinephrine acts on alpha and beta adrenergic receptors. The effects most people feel are a faster heart rate and stronger heart contractions, and it also constricts some blood vessels, dilates the pupils and triggers renin release in the kidney.3 The adrenal medulla releases it in the first seconds of a perceived threat.1

  • TriggersSudden threat, hard exertion, public speaking and stimulants. In a randomized crossover trial, caffeine was linked to repeated daily blood pressure elevations and higher stress reactivity, with epinephrine among the measures tracked.9
  • CountersSlow breathing is the most studied voluntary lever. A meta-analysis examined its effect on heart rate variability, and a separate meta-analysis of randomized trials asked whether breathwork lowers self-reported stress.1011
  • ManageTime caffeine away from known stress peaks such as a presentation, and use slow breathing before and during them.
  • Weak pointNeither review measured circulating adrenaline. Heart rate variability and survey scores are proxies, so I cannot claim breathing lowers adrenaline itself.

03 · Cortisol: a rhythm, not a villain

Cortisol follows a daily rhythm with a rise after waking, known as the cortisol awakening response. Researchers measure it with saliva samples that must be timed from the moment of waking, which is hard to do correctly at home.4 Chronic stress can push the HPA axis out of balance, and reviews describe direct effects on the heart and blood vessels through glucocorticoid signaling.2

  • TriggersPsychosocial stress, and caffeine. In a double-blind crossover trial of 96 adults, challenge doses raised cortisol across the day after five days off caffeine.12
  • Not a triggerAcute sleep deprivation. A meta-analysis of 24 studies found no significant overall difference in cortisol, which cuts against the common claim that one bad night spikes it.13
  • CountersA meta-analysis of randomized trials in non-patient groups found mindfulness and meditation (g = 0.345) and relaxation (g = 0.347) changed cortisol most among stress management types.14 A review of ten studies found moderate-certainty evidence that physical activity lowered cortisol (SMD -0.37).15 A separate meta-analysis found meditation lowered physiological stress markers including blood pressure, heart rate and cortisol.16
  • ManagePick one daily practice from the list above and keep it for eight weeks before judging it. Keep caffeine off the morning if you are trying to read your own stress level.
  • Weak pointEffects are small to moderate and vary by measurement method. Lower cortisol is not automatically better, because a healthy rhythm includes a morning peak.4

04 · Oxytocin: context decides the effect

Oxytocin is a hypothalamic neuropeptide linked to social cognition and behavior. Most human experiments give it as a nasal spray, and the review literature concludes that the effects on social behavior are more complex than early reports suggested.5 That caution applies to the popular claim that a hug floods the brain with oxytocin.

  • TriggersSocial support and touch are the commonly cited triggers. Men given intranasal oxytocin plus social support secreted less cortisol under a standard stress test than men given support alone.17 In a 159-person randomized trial, self-soothing touch and being hugged both lowered post-stress cortisol, while heart rate and self-reported stress did not differ.18
  • CountersOxytocin does not need counteracting. The relevant question is what blunts it, and the research I found does not answer that.
  • ManageSchedule real contact before a known stressor, such as time with a partner or friend. The evidence supports a cortisol effect, not an oxytocin effect.
  • Weak pointThe touch trial did not measure oxytocin, and a partner-embrace study found a cortisol benefit in women and none in men.1918 Treat “hugs release oxytocin and lower stress” as plausible and unproven.

05 · Serotonin: two separate pools

About 95% of the body’s serotonin is made in the gut by enterochromaffin cells.6 That serotonin does not enter the brain, because it cannot cross the blood-brain barrier.7 Brain serotonin comes from its own neurons. My reading of those two facts is that food or gut serotonin does not directly top up the mood supply, though the gut still affects the brain through other routes.

  • TriggersI found no consumer-level lever with strong human evidence for raising brain serotonin directly. That gap is itself a finding.
  • The depression debateA 2022 umbrella review found no consistent evidence that low serotonin causes depression.20 It did not test whether antidepressants work, and other researchers dispute how it weighed the evidence.20
  • ManageTreat persistent low mood as a medical question and take it to a clinician. A supplement or diet change is not a tested substitute.
  • Weak pointThe most-quoted claims on this topic come from blogs and wellness sites. I excluded them and relied on reviews and abstracts.

06 · Dopamine: motivation more than pleasure

A review of dopamine neuron recordings describes at least two kinds of signal. Some neurons encode motivational value, which supports seeking and learning. Others encode motivational salience, which supports orienting and general drive.8 The same review sets the reward-learning theory against the incentive-signal theory, which is why “pleasure chemical” overstates what the evidence supports.8

  • TriggersNovelty and rewards that arrive unexpectedly, based on the reward-prediction account the review covers.8
  • ExerciseThe evidence conflicts. An early PET study found no detectable striatal dopamine change from aerobic exercise, while a 2024 study linked moderate exercise, dopamine measures and faster reaction time in a small correlational design.2122
  • ManageThis is my judgment, not a tested protocol: limit variable-reward loops such as feeds and notifications during focus work, and place rewards after effort.
  • Weak point“Dopamine detox” advice has no study in this pull behind it, so I do not endorse it.

07 · What you can and cannot measure

The market for home tests runs ahead of the science. The only item on this list with a published collection standard is the cortisol awakening response.

Whether each chemical can be measured, and how
ChemicalWhat can be measuredWhat cannot
AdrenalineClinical lab tests for catecholamines. At home, heart rate and heart rate variability act as proxies.1Real-time levels outside a clinic.
CortisolTimed saliva samples for the awakening response, collected by a strict schedule.4Sloppy sampling gives invalid curves.4
OxytocinResearch assays only.No home test links blood levels to brain effects.
SerotoninBlood or gut measures reflect the peripheral pool.67Brain levels are not reachable by blood or urine tests.
DopamineBrain imaging in research settings, such as PET.21No home test.

The oxytocin row and the home-test claims are my judgment from the absence of validated consumer assays in the sources I pulled. They are not cited findings.

08 · What to try, ranked by evidence

This ranking reflects the sources above. It ranks evidence quality, not how well a lever works for you.

Practical levers ranked by strength of the evidence found
LeverWhat it targetsEvidence foundStrength
Mindfulness or relaxation practiceCortisolLargest effects among stress management types in a randomized-trial meta-analysis14Moderate
Regular physical activityCortisolTen studies, moderate certainty, SMD -0.3715Moderate
Slow breathingHeart rate variability, self-reported stressMeta-analyses on HRV and on subjective stress1011Low to moderate. Proxy outcomes.
Caffeine timingCortisol, epinephrine, blood pressureCrossover trials show higher cortisol and stress reactivity129Moderate
Social contact before stressCortisolTrials show lower cortisol, with sex differences and no oxytocin measure1819Low
Gut or diet changes for serotoninBrain serotoninNo direct evidence in this pull7None found

This is a reading of published research, not medical advice.

Palpitations, persistent low mood or unexplained fatigue belong with a clinician, and none of the levers above substitute for one.

09 · Where this piece is weakest

  1. Five blocked pages (including several PubMed and PMC records) forced me to rely on search excerpts. Those sources carry flags in the footnotes.
  2. Most cortisol effect sizes come from healthy adults in meta-analyses with mixed methods, so they may not transfer to a person with a clinical condition.
  3. The dopamine exercise evidence rests on two small studies that disagree.
  4. Several manage lines are my judgment and say so. I did not test them.
  5. Studies on touch and social support measured cortisol, not oxytocin, so the oxytocin story is a hypothesis.

Sources and notes

  1. Tank, A. William, and Dona Lee Wong. 2015, January. “Peripheral and central effects of circulating catecholamines.” Comprehensive Physiology. p. 1 to 15. https://pubmed.ncbi.nlm.nih.gov/25589262/ Back

    Supports the point that epinephrine and norepinephrine from the adrenal medulla drive the rapid fight-or-flight response.

  2. Review article. 2017. Title and authors unverified. International Journal of Molecular Sciences (volume 18, article 2150). p. n/a. https://pmc.ncbi.nlm.nih.gov/articles/PMC5666832 Back

    Reviews stress effects on the cardiovascular system through HPA axis activation and glucocorticoid signaling in blood vessels and the heart. Title and authors were not captured, so bibliographic fields are omitted pending a human check.

  3. StatPearls Publishing. Year unverified. “Epinephrine.” StatPearls, NCBI Bookshelf. p. n/a. https://www.ncbi.nlm.nih.gov/books/NBK482160/ Back

    Describes epinephrine action on alpha and beta adrenergic receptors, including higher heart rate, contractility and renin release. The page blocked automated access, so I used the search excerpt and it needs a human check.

  4. Stalder, Tobias, et al. 2016, January. “Assessment of the cortisol awakening response: Expert consensus guidelines.” Psychoneuroendocrinology. p. 414 to 432. https://pubmed.ncbi.nlm.nih.gov/26563991/ Back

    States that CAR is usually collected as home saliva samples and is valid only if participants follow a timed schedule starting at the moment of waking.

  5. Quintana, Daniel S., et al. 2021, January. “Advances in the field of intranasal oxytocin research: lessons learned and future directions for clinical research.” Molecular Psychiatry. p. n/a. https://pmc.ncbi.nlm.nih.gov/articles/PMC7815514 Back

    Notes that human oxytocin research mostly uses intranasal protocols and that effects on social behavior are more complex than first thought. The page blocked automated access, so the author and date fields need a human check.

  6. Gut-derived serotonin review. 2022, April. “Gut-derived serotonin and its emerging roles in immune function, inflammation, metabolism and the gut-brain axis.” Current Opinion in Endocrinology, Diabetes and Obesity. p. 177 to 182. https://pubmed.ncbi.nlm.nih.gov/35197425/ Back

    States that peripheral serotonin makes up about 95% of the total and is largely made by gut enterochromaffin cells. Authors are omitted because I did not verify them.

  7. Review article. Year, title and authors unverified. Peer-reviewed review on serotonin transport. p. n/a. https://pmc.ncbi.nlm.nih.gov/articles/PMC9627707 Back

    States that peripherally synthesized serotonin does not enter the brain because it cannot cross the blood-brain barrier. Taken from a search excerpt, so title and authors need a human check.

  8. Bromberg-Martin, Ethan S., Masayuki Matsumoto, and Okihide Hikosaka. 2010. “Dopamine in motivational control: rewarding, aversive, and alerting.” Neuron. p. n/a. https://pmc.ncbi.nlm.nih.gov/articles/PMC3032992/ Back

    Review of how some dopamine neurons encode motivational value and others encode salience, and how reward prediction and incentive theories differ.

  9. Lane, James D., et al. 2002, July. “Caffeine affects cardiovascular and neuroendocrine activation at work and home.” Psychosomatic Medicine. p. 595 to 603. https://pubmed.ncbi.nlm.nih.gov/12140349/ Back

    Randomized crossover trial. The abstract links caffeine to repeated daily blood pressure elevations and higher stress reactivity, and lists epinephrine, norepinephrine, cortisol and heart rate among the measures. Author list is from memory and needs a check.

  10. Laborde, Sylvain, et al. 2022. “Effects of voluntary slow breathing on heart rate and heart rate variability: A systematic review and a meta-analysis.” Neuroscience and Biobehavioral Reviews. p. n/a. https://fis.dshs-koeln.de/de/publications/effects-of-voluntary-slow-breathing-on-heart-rate-and-heart-rate-/ Back

    Reviews slow breathing and heart rate variability. I could not open the full text, so I cite it only for the research question.

  11. Fincham, Guy W., et al. 2023, January. “Effect of breathwork on stress and mental health: A meta-analysis of randomised-controlled trials.” Scientific Reports. p. 432. https://pmc.ncbi.nlm.nih.gov/articles/PMC9828383/ Back

    Meta-analysis built around whether breathwork lowers self-reported stress compared with non-breathwork controls. The measure is subjective stress, not a hormone level.

  12. Lovallo, William R., et al. 2005, September. “Caffeine stimulation of cortisol secretion across the waking hours in relation to caffeine intake levels.” Psychosomatic Medicine. p. 734 to 739. https://pubmed.ncbi.nlm.nih.gov/16204431/ Back

    Double-blind crossover trial in 96 adults. After five days off caffeine, challenge doses raised cortisol across the day.

  13. Endocrine Journal. 2024, May. “The effect of acute sleep deprivation on cortisol level: a systematic review and meta-analysis.” Endocrine Journal. p. n/a. https://pubmed.ncbi.nlm.nih.gov/38777757/ Back

    Pooled 24 studies and found no significant overall cortisol difference after acute sleep deprivation. Authors omitted because I did not verify them.

  14. Rogerson, Olivia, et al. 2023. “Effectiveness of stress management interventions to change cortisol levels: a systematic review and meta-analysis.” Psychoneuroendocrinology. p. n/a. https://eprints.whiterose.ac.uk/205257/ Back

    Random-effects meta-analysis of randomized trials in non-patient groups. Mindfulness and meditation (g = 0.345) and relaxation (g = 0.347) had the largest effects on cortisol.

  15. Physical activity and cortisol review. 2022, September. “The effects of physical activity on cortisol and sleep: A systematic review and meta-analysis.” Psychoneuroendocrinology. p. n/a. https://pubmed.ncbi.nlm.nih.gov/35777076/ Back

    Ten studies. Moderate-certainty evidence that physical activity lowered cortisol (SMD -0.37, 95% CI -0.52 to -0.21) and low-certainty evidence for sleep. Authors omitted because I did not verify them.

  16. Pascoe, Michaela C., et al. 2017, December. “Mindfulness mediates the physiological markers of stress: Systematic review and meta-analysis.” Journal of Psychiatric Research. p. 156 to 178. https://pubmed.ncbi.nlm.nih.gov/28863392/ Back

    Meta-analysis of randomized trials against active controls. Concludes meditation practice decreased physiological markers of stress, with blood pressure, heart rate and cortisol among the outcomes indexed.

  17. Heinrichs, Markus, et al. 2003. “Social support and oxytocin interact to suppress cortisol and subjective responses to psychosocial stress.” Biological Psychiatry. p. n/a. https://pubmed.ncbi.nlm.nih.gov/14675803/ Back

    Men given intranasal oxytocin plus social support secreted less cortisol under the Trier Social Stress Test than men given support alone. The page blocked automated access, so I relied on the way a later paper describes it. Needs a human check.

  18. Packheiser, Julian, et al. 2022. “Self-soothing touch and being hugged reduce cortisol responses to stress: A randomized controlled trial on stress, physical touch, and social identity.” Comprehensive Psychoneuroendocrinology. p. n/a. https://pubmed.ncbi.nlm.nih.gov/35757667/ Back

    159 participants. Both touch conditions had lower post-stress cortisol than controls. Heart rate and self-reported stress did not differ. Oxytocin was not measured. Author list is from memory and needs a check.

  19. Berretz, Gesa, et al. 2022, May. “Romantic partner embraces reduce cortisol release after acute stress induction in women but not in men.” PLOS ONE. p. n/a. https://pmc.ncbi.nlm.nih.gov/articles/PMC9116618 Back

    Women who embraced a partner before the stressor had a smaller cortisol response than controls. No effect appeared in men. The page blocked automated access, so I relied on the search excerpt. Needs a human check.

  20. Moncrieff, Joanna, et al. 2022, July. “The serotonin theory of depression: a systematic umbrella review of the evidence.” Molecular Psychiatry. p. n/a. https://discovery.ucl.ac.uk/10153006/3/Horowitz_s41380-022-01661-0.pdf Back

    Umbrella review of the evidence for a link between serotonin and depression. It does not test whether antidepressants work, and its conclusions are disputed by other researchers.

  21. Wang, Gene-Jack, et al. 2000. “PET Studies of the Effects of Aerobic Exercise on Human Striatal Dopamine Release.” Journal of Nuclear Medicine. p. 1352 to 1356. https://jnm.snmjournals.org/content/jnumed/41/8/1352.full-text.pdf Back

    Small PET study in which aerobic exercise did not produce a detectable change in striatal dopamine release. I did not open the full text, so the null result is from my notes and needs a check.

  22. Ando, et al. 2024, January. “The neuromodulatory role of dopamine in improved reaction time by acute cardiovascular exercise.” Journal of Physiology. p. n/a. https://eprints.chi.ac.uk/id/eprint/7325/ Back

    Small PET study linking moderate exercise, dopamine measures and faster reaction time. The design is correlational. Author names and journal need a human check.

← Back to Fitness