Massage, Cortisol and Stress: Why The Story Needs To Change

For years, massage has been sold with a neat promise:

‘Massage reduces cortisol by about 30 per cent.’

It sounds scientific and powerful. It is also too simple and keeps dragging massage into a narrow biochemical story that does not match what really happens in practice.

Modern research suggests something different. Massage is not a cortisol treatment. It is a form of supportive, skilful touch that can act as a stress buffer. People usually report feeling calmer, less wound up and more comfortable in their bodies after a session. Any change in cortisol sits downstream from that broader shift in the nervous system and it depends heavily on the person’s own life, stress history and daily rhythm.

If we want to stay honest as a profession, we need to stop centring massage’s value on one hormone.

Where the ‘massage reduces cortisol’ story came from

In 2005, Tiffany Field and colleagues published a narrative review that pulled together numerous massage studies and reported that:

• Cortisol fell by about 31 per cent after massage
• Serotonin and dopamine rose by about 28–31 per cent

Those numbers were eye catching. They spread into training manuals, trade magazines and marketing copy. Massage became framed as something that directly lowered stress hormones and raised feel good chemicals.

But that review had serious limits:

• Many included trials were small and short
• Conditions, populations and massage protocols were very mixed
• Controls were weak or missing
• Cortisol was usually measured once before and once after a session, with poor control for time of day, sleep, food, medication or pain

Cortisol does not sit still. It follows a daily curve, peaks in the morning, falls through the day and responds to life events, not just to treatment. A single pre and post sample is just a snapshot on that curve. A drop of 20 or 30 per cent might be nothing more than normal progression through the day.

So the bold message that ‘massage reduces cortisol by 31 per cent’ came from optimistic data and a very narrow focus on one marker.

Moyer 2011: correcting the numbers but keeping the same tunnel vision

Christopher Moyer and colleagues pushed back against the enthusiastic cortisol story.

In 2004, their meta analysis of randomised massage trials showed that single sessions produced clear changes in state anxiety, blood pressure and heart rate, but not cortisol.

In 2011, they went further with a quantitative review titled ‘Does massage therapy reduce cortisol?’. They used stricter methods, focusing on between group differences and higher quality trials.

They found:

• Small and often non significant effects of massage on cortisol compared with controls
• Some moderate effects in a small number of child studies, with clear risk of publication bias
• No convincing link between cortisol changes and the much stronger effects seen on anxiety, depression and pain

Moyer was absolutely right to knock down the idea that massage’s benefits are driven by big hormone shifts. However, his review still treated cortisol as the central question. In effect, Field gave us a story of ‘massage clearly reduces cortisol’, and Moyer replied with ‘massage probably does not significantly reduce cortisol’.

Both views are myopic. Both imply that cortisol is where we should be looking for the main therapeutic action. Neither reflects what clients actually report, or how we now understand touch and stress.

• Were especially effective in regulating cortisol and increasing weight in newborns
• Reduced pain, depression and anxiety with moderate effect sizes in children and adults
• Had stronger mental health effects when touch was human rather than robotic

This work does not say that touch always lowers cortisol. It shows that touch has measurable effects on stress related systems, but again it highlights that:

• Context matters (medical setting, home, lab stress test)
• Who is doing the touching matters (parent, partner, professional, device)
• Baseline state matters (healthy, acutely stressed, chronically stressed, premature baby)

Other recent studies underline the same point from different angles:

• Dreisoerner and colleagues found that self soothing touch and being hugged reduced cortisol responses to a standard lab stress test, but changes in heart rate and self reported stress were less clear.
• Berretz and colleagues showed that women who embraced a romantic partner before a stress task had a smaller cortisol spike than women who did not, despite little difference in blood pressure or mood scores.
• Romney and colleagues reported associations between frequent hugs and a healthier cortisol awakening response the next day.
• A 2025 review found that gentle touch in newborns supports self regulation and sometimes shifts cortisol, but in different directions depending on baseline stress and the design of the intervention.

Taken together, these findings say:

Touch clearly influences stress systems, including cortisol, but the direction and size of change depend on the person and context. Cortisol is part of the picture, not the lead actor.

Why cortisol on its own is the wrong target

Cortisol sits inside the hypothalamic pituitary adrenal, HPA, axis. That system is shaped by each person’s:

• Sleep pattern
• Ongoing life stress, caring roles and financial pressure
• Pain, injury and illness
• Medication
• Past history of stress, burnout or trauma
• Diet, caffeine and alcohol

No two people have the same starting point.

That means:

• A single person’s cortisol can move up or down on different days, even without treatment
• A group average change tells you very little about what happened to any individual
• A small rise in cortisol is not automatically harmful, and a small drop is not automatically helpful

So when Field highlights a 31 per cent mean drop and Moyer highlights tiny mean effects, both are still locked on the same limited question, ‘What happens to this one hormone?’

Massage and touch ask bigger questions:

• How does this person feel before and after
• What happens to their sleep, pain, energy and ability to cope over weeks and months
• Does regular supportive touch act as a stress buffer, easing the workload on their stress systems, even if hormone levels sit within their own normal range

Those are the outcomes that clients care about and that therapists actually see.

Stress buffers, not hormone hacks

A more useful way to frame massage is as a stress buffer rather than a hormone hack.

Massage is more than touch alone. It is the entire therapeutic interaction wrapped around the touch:

• The welcome, first contact and ease of booking
• The feeling of being expected and not rushed
• The room, temperature, lighting, music and privacy
• The therapist’s demeanour, kindness and consistency
• The sense that the therapist is approachable, trustworthy and on the client’s side
• The experience of feeling listened to, believed and safe

Massage strokes sit inside that bigger frame. During a therapy the person usually experiences:

• Predictable, safe, warm touch
• A quieter environment with fewer demands
• Time where they are not performing, caring or working
• A sense of being looked after rather than having to look after others

All of this is the therapeutic interaction. It is the developing therapeutic relationship that allows the nervous system to downshift. Sympathetic drive, constant alertness and readiness to react, eases off. Parasympathetic processes that support rest, digestion and repair can come forward.

Cortisol and other stress hormones sit downstream from that broader shift. If the nervous system spends less time in a high threat, high alert state, the HPA axis is likely to have less work to do. For some people that will show up as smaller cortisol spikes or a smoother daily curve. For others, the hormone numbers will sit within their usual range, but they still feel and function better.

Either way, the story is not:

‘Massage directly reduces cortisol and that is why people feel better.’

It is closer to:

‘Massage and the wider therapeutic interaction can act as a stress buffer. They help the nervous system step out of constant stress mode, which reduces the load on the body’s stress systems. Hormones like cortisol then adjust within each person’s own normal range.’

That is a very different message. It does not promise that we can drive cortisol up or down to a specific target. It places hormone change, if it happens at all, as a side effect of broader regulation, not as the primary goal.

How to talk about cortisol without getting trapped by it

To get out of the cortisol rabbit hole, it helps to tidy up our language.

What to drop

Let go of phrases like:

• ‘Massage reduces cortisol by 31 per cent.’
• ‘Massage flushes cortisol out of your system.’
• ‘Massage rebalances your hormones.’

These are not supported by current evidence and invite complaints if challenged.

What to say instead

For clients and the public:

• ‘Massage offers a calm, safe space where your body can slow down from everyday stress.’
• ‘Regular massage can be one useful part of managing stress and tension, alongside sleep, movement and other self care.’
• ‘Hormones like cortisol naturally go up and down. Massage is not a hormone treatment, but people often notice they feel calmer, less tense and more able to cope afterwards.’

That framing keeps us out of the hormone promise business and lets us focus on outcomes that are real, observable and well supported.

The bottom line

The profession has spent too long arguing over whether massage does or does not reduce cortisol. Both positions are too narrow.

The evidence now points to a broader and more useful conclusion:

• Massage is not a cortisol treatment.
• Massage, including the whole therapeutic interaction around it, can act as a stress buffer, helping many people feel less anxious, less tense and more at ease in their bodies.
• Hormones like cortisol naturally fluctuate and are different for each person. Any changes are secondary to how the nervous system and the person as a whole respond, not the main marker of therapeutic success.

What matters most is not a percentage shift in a lab test, but the client getting off the couch and genuinely feeling that life is a little easier to handle.

References

Field, T., Hernandez-Reif, M., Diego, M., Schanberg, S. and Kuhn, C. (2005) ‘Cortisol decreases and serotonin and dopamine increase following massage therapy’, International Journal of Neuroscience, 115(10), pp. 1397–1413. Available at: https://doi.org/10.1080/00207450590956459 PubMed+1

Moyer, C.A., Rounds, J. and Hannum, J.W. (2004) ‘A meta-analysis of massage therapy research’, Psychological Bulletin, 130(1), pp. 3–18. Available at: https://doi.org/10.1037/0033-2909.130.1.3 PubMed+1

Moyer, C.A., Seefeldt, L., Mann, E.S. and Jackley, L.M. (2011) ‘Does massage therapy reduce cortisol? A comprehensive quantitative review’, Journal of Bodywork and Movement Therapies, 15(1), pp. 3–14. Available at: https://doi.org/10.1016/j.jbmt.2010.06.001 PubMed+2ScienceDirect+2

Packheiser, J., Hartmann, H., Friedrich, M., Koerner, A. and Ocklenburg, S. (2024) ‘A systematic review and multivariate meta-analysis of the physical and mental health benefits of touch interventions’, Nature Human Behaviour, 8, 1841–1854. Available at: https://doi.org/10.1038/s41562-024-01841-8 Helena Hartmann+3Nature+3PubMed+3

Dreisoerner, A., Junker, N.M., Schlotz, W., Heimrich, J., Bloemeke, S., Ditzen, B. and van Dick, R. (2021) ‘Self-soothing touch and being hugged reduce cortisol responses to stress: A randomised controlled trial on stress, physical touch, and social identity’, Comprehensive Psychoneuroendocrinology, 8, 100091. Available at: https://doi.org/10.1016/j.cpnec.2021.100091 ScienceDirect+2PubMed+2

Berretz, G., Wolf, O.T., Güntürkün, O. and Ocklenburg, S. (2022) ‘Romantic partner embraces reduce cortisol release after acute stress induction in women but not in men’, PLOS ONE, 17(5), e0266887. Available at: https://doi.org/10.1371/journal.pone.0266887 PLOS+2PubMed+2

Romney, C.E., Arroyo, A.C., Robles, T.F. and Zawadzki, M.J. (2023) ‘Hugs and cortisol awakening response the next day: An ecological momentary assessment study’, International Journal of Environmental Research and Public Health, 20(7), 5340. Available at: https://doi.org/10.3390/ijerph20075340 PubMed+2MDPI+2

Ionio, C., Ciuffo, G. and Landoni, M. (2021) ‘Parent–infant skin-to-skin contact and stress regulation: A systematic review of the literature’, International Journal of Environmental Research and Public Health, 18(9), 4695. Available at: https://doi.org/10.3390/ijerph18094695

Valori, I., Merkel, J.L., Furlan, G., Grossmann, T. and Fairhurst, M.T. (2025) ‘Touch facilitates newborns’ self-regulation: Systematic review of multidimensional arousal outcomes’, Neuroscience and Biobehavioral Reviews, 178, 106353. Available at: https://doi.org/10.1016/j.neubiorev.2025.106353