What massage actually does to your brain: the evidence, graded

Massage does reach your brain, through a real and well-mapped nerve pathway, and that part of the science is solid. The chemistry claims are a different story. Lower cortisol, more serotonin, more dopamine, a rush of oxytocin: the evidence behind each is much weaker than most wellness websites suggest.

We graded eight common claims about massage and the brain by the strength of the evidence behind them, the way a strict teacher would. Some passed with honours. Others barely scraped through, and one failed outright.

The report card

Grades are SEDINA's editorial judgement of evidence strength, not a formal scoring system. The reasoning for each is below, with links to the studies.

Grade A: the pleasant-touch pathway

Your skin has nerves built for pleasant touch (A). C-tactile fibres in hairy skin fire most at slow strokes of about 1 to 10 cm per second, and the speeds they prefer are the speeds people rate as most pleasant (Löken et al., 2009). Their signal travels mainly to the posterior insula, the brain region that tracks how your body feels. Our companion guide, your skin has a speed limit for pleasure, covers this in depth.

Your brain responds more to a moving human hand (A−). In a 2012 brain-imaging study from Umeå University in Sweden, volunteers were touched in four ways: by a bare human hand or a rubber glove, each either moving or still. Pressure and speed were held constant, at a slow 1.5 cm per second. The moving human hand was rated most pleasant, and it produced the strongest activation in the pregenual anterior cingulate cortex, a region linked to reward and emotion (Lindgren et al., 2012).

The practical point: the brain distinguishes how it is touched, not just whether it is. The same slow stroke through a rubber glove was not processed the same way as skin on skin. The touch tested was modelled on touch massage, a gentle technique used in clinical care, which puts this study closer to the massage table than most lab work. We give it an A− because it rests mainly on one study, though it fits a wider body of imaging research.

Grade B: pain relief, modest and short-term

You already know this one. When you bump your elbow, you rub it, and it hurts a little less.

The classic explanation dates to 1965, when Ronald Melzack and Patrick Wall proposed the gate-control theory of pain. Pressure and rubbing activate fast touch nerves. Those signals compete with slower pain signals at a "gate" in the spinal cord, turning down how much pain reaches the brain (Melzack & Wall, 1965). The details of the original model have been revised since, but its core idea, that touch can damp down pain signals, is widely accepted.

The grade is B, not A, because the clinical results are modest. A Cochrane review of 25 trials of massage for low-back pain found that massage reduced pain more than inactive comparisons such as sham treatment, but only at short-term follow-up, and it rated the evidence as low to very low quality (Furlan et al., 2015). A 2004 meta-analysis found that single sessions did not reliably lower pain ratings taken straight afterwards, while a course of sessions reduced pain measured later (Moyer et al., 2004).

Massage can take the edge off for a while. It is not a treatment for the cause of pain. If pain is persistent or unexplained, see a doctor or a registered fisioterapeuta; massage or physiotherapy explains how to choose.

Grade C: promising but unproven

"Rest and digest" (C). Your nervous system has an accelerator (the stress response) and a brake (the parasympathetic, or "rest and digest", system). Researchers track the brake through heart-rate variability (HRV). In a 2020 laboratory trial, 60 healthy women received a vagus-nerve massage, a soft shoulder massage or simply rested. Felt relaxation and the vagal component of HRV rose in all three groups, and both massages raised HRV more than resting did (Meier et al., 2020).

The results are not consistent, though. A small 2025 placebo-controlled trial in 15 women found no significant difference in HRV or blood pressure between a two-minute manual massage, a sham treatment and plain rest (Monteiro et al., 2025). An older meta-analysis found that single sessions lowered heart rate, blood pressure and self-reported anxiety (Moyer et al., 2004). Most studies are small and unblinded, so we grade this C.

The brain's own opioids (C). In a 2016 study, 18 men had PET brain scans while their partners caressed them in a non-sexual way, and again while lying alone. Social touch felt pleasant and changed the availability of μ-opioid receptors, part of the brain's own pain-relief and reward system, in regions including the thalamus, striatum, cingulate and insula (Nummenmaa et al., 2016).

This is a direct measure inside the brain, which is rare and valuable. But it involved 18 men touched by their own partners, not massage from a professional, and it has not been widely replicated. The receptor change is also not simple to read: the authors describe it as modulation, and the study does not show that touch "releases natural painkillers". It is an interesting clue, not a conclusion.

Grades D and F: cortisol, oxytocin, serotonin and dopamine

Cortisol (D). You will often read that massage lowers the stress hormone cortisol by around 31%. That figure comes from a 2005 review that averaged before-and-after changes within massage groups, across studies mostly run by the same research group (Field et al., 2005).

In 2011, an independent team led by Christopher Moyer re-analysed the research using conventional between-group comparisons, which set each massage group against a control group. The effects were almost all small and not statistically significant. Their conclusion: the effect of massage on cortisol is generally very small, and in most studies it cannot be told apart from zero (Moyer et al., 2011). The one exception, a larger effect in children after several sessions, rested on just three studies. An earlier 2004 meta-analysis, also led by Moyer, had found no reliable cortisol effect after single sessions either. Massage and the immune system goes through the cortisol evidence in more detail.

Oxytocin (D). Some studies report oxytocin rising in the blood after massage. Others do not, and even a positive 2012 trial noted that earlier results had been inconsistent (Morhenn et al., 2012). The bigger problem is measurement. Widely used commercial tests for oxytocin in blood have been shown to be unreliable on unprocessed samples and to pick up molecules other than oxytocin (McCullough et al., 2013). And at rest, blood levels do not track brain levels: a 2017 meta-analysis found no association between the two under normal conditions (Valstad et al., 2017). Until better methods are used, treat "massage releases the love hormone" as a slogan, not a finding.

Serotonin and dopamine (F). The same 2005 review reported average increases of 28% in serotonin and 31% in dopamine. In those studies, both were measured in urine. Urine is a poor proxy for brain chemistry. Much of the body's serotonin is made in the gut (Yano et al., 2015), and serotonin made in the body does not cross into the brain. Dopamine is no better: much of the dopamine in urine is made by the kidneys themselves. Wellness sites still repeat these numbers as if they described your mood. They don't. This claim fails.

Why the evidence is weaker than you think

Scientists have a problem: you can't fake a massage.

Drug trials compare a real pill with an identical sugar pill, so neither patient nor doctor knows who got what. Massage has no convincing equivalent. People know when they are being massaged, and the therapist always knows. That makes it very hard to separate the effect of touch from the effect of expecting to feel better, resting for an hour or simply being cared for. The Cochrane review above names exactly this problem: the difficulty of blinding participants, therapists and outcome measurement was the most common source of bias.

Other weaknesses run across the field:

• Small samples. Many studies involve a few dozen people. The opioid study had 18; the 2025 heart-rate trial had 15.
• Unblinded measurement. Participants, and often researchers, know which group is which.
• Concentrated sources. A large share of early positive findings came from a single research group, which makes independent replication more important.
• Indirect measures. Urine, saliva and blood are easier to sample than the brain, but they often measure the wrong thing.

None of this means massage does nothing. It means the honest answer to "what does it do?" is narrower than the marketing. The wellness tax looks at what that marketing costs you.

What this means for you

You don't need a hormone chart to justify a massage. The best-supported reasons are simpler: it feels good, your nervous system is wired to register that, and skin-to-skin touch is processed differently from the same stroke through a glove.

• Book it for enjoyment, not for a cure. Pleasure and an hour of calm are good enough reasons.
• Be sceptical of numbers. If a website promises "31% less cortisol" or a serotonin boost, you now know where those figures come from.
• Choose the person, not the promise. A therapist who listens, works unhurriedly and adapts pressure to you matters more than a menu of claims. How to read massage reviews helps you spot one.
• See a professional for health problems. For persistent pain or a medical condition, start with a doctor or a registered fisioterapeuta.

Frequently asked questions

What does massage actually do to your brain?
The best-supported effect is a real nerve pathway. Slow, gentle stroking activates C-tactile fibres in hairy skin, whose signal reaches the posterior insula, and moving skin-to-skin touch activates reward-linked areas such as the pregenual anterior cingulate cortex. Massage can also ease pain modestly and for a while, and single sessions lower heart rate, blood pressure and anxiety in the short term. Claims about cortisol, oxytocin, serotonin and dopamine are much weaker.

Does massage lower cortisol?
Barely, if at all. The popular figure of a 31% drop comes from a 2005 review that averaged before-and-after changes within massage groups. A 2011 meta-analysis that compared massage groups with control groups found the effect on cortisol was generally very small and, in most studies, could not be told apart from zero.

Does massage increase serotonin and dopamine?
There is no good evidence that it raises them in the brain. The often-quoted increases of 28% for serotonin and 31% for dopamine were measured in urine, which says little about brain chemistry: much of the body's serotonin is made in the gut, serotonin made in the body does not cross into the brain, and much of the dopamine in urine is made by the kidneys.

Does massage release oxytocin?
Some studies report higher oxytocin in the blood after massage and others do not. Widely used blood tests for oxytocin have been shown to be unreliable, and under normal resting conditions blood levels do not track levels in the brain, so these results cannot show what happens in your head.

Is massage good for stress and anxiety?
Modestly, and mostly in the short term. A meta-analysis of 37 randomised studies found that a single massage reduced momentary anxiety, blood pressure and heart rate, and that a course of sessions reduced anxiety and depression scores. Massage is not a substitute for care: persistent anxiety or low mood should be raised with a doctor.

Sources

• Field T., Hernandez-Reif M., Diego M., Schanberg S., Kuhn C. (2005). Cortisol decreases and serotonin and dopamine increase following massage therapy. Int J Neurosci 115(10):1397–1413.
• Furlan A.D., Giraldo M., Baskwill A., Irvin E., Imamura M. (2015). Massage for low-back pain. Cochrane Database Syst Rev (9):CD001929.
• Lindgren L., Westling G., Brulin C., Lehtipalo S., Andersson M., Nyberg L. (2012). Pleasant human touch is represented in pregenual anterior cingulate cortex. NeuroImage 59(4):3427–3432.
• Löken L.S., Wessberg J., Morrison I., McGlone F., Olausson H. (2009). Coding of pleasant touch by unmyelinated afferents in humans. Nat Neurosci 12(5):547–548.
• McCullough M.E., Churchland P.S., Mendez A.J. (2013). Problems with measuring peripheral oxytocin: can the data on oxytocin and human behavior be trusted?. Neurosci Biobehav Rev 37(8):1485–1492.
• Meier M., Unternaehrer E., Dimitroff S.J. et al. (2020). Standardized massage interventions as protocols for the induction of psychophysiological relaxation in the laboratory: a block randomized, controlled trial. Sci Rep 10:14774.
• Melzack R., Wall P.D. (1965). Pain mechanisms: a new theory. Science 150(3699):971–979.
• Monteiro E.R., Pescatello L.S. et al. (2025). Upper cervical manipulation and manual massage do not modulate sympatho-vagal balance or blood pressure in women: a randomized, placebo-controlled clinical trial. Healthcare (Basel) 13(20):2554.
• Morhenn V., Beavin L.E., Zak P.J. (2012). Massage increases oxytocin and reduces adrenocorticotropin hormone in humans. Altern Ther Health Med 18(6):11–18.
• Moyer C.A., Rounds J., Hannum J.W. (2004). A meta-analysis of massage therapy research. Psychol Bull 130(1):3–18.
• Moyer C.A., Seefeldt L., Mann E.S., Jackley L.M. (2011). Does massage therapy reduce cortisol? A comprehensive quantitative review. J Bodyw Mov Ther 15(1):3–14.
• Nummenmaa L., Tuominen L., Dunbar R. et al. (2016). Social touch modulates endogenous μ-opioid system activity in humans. NeuroImage 138:242–247.
• Valstad M., Alvares G.A., Egknud M. et al. (2017). The correlation between central and peripheral oxytocin concentrations: a systematic review and meta-analysis. Neurosci Biobehav Rev 78:117–124.
• Yano J.M., Yu K., Donaldson G.P. et al. (2015). Indigenous bacteria from the gut microbiota regulate host serotonin biosynthesis. Cell 161(2):264–276.

SEDINA is free, takes no commission and sells no placement. Answer three questions and get three verified venues, three, deliberately. 195 of the venues SEDINA lists offer relaxation massage. SEDINA covers Barcelona (81) · Madrid (96) · Málaga (42) · Valencia (64) · Seville (24).