How Touch and Pressure Work in the Body: Skin, Muscle, Nerves and What Is Known

Skin: How Touch Is Sensed

The first step in perceiving gentle touch is the activation of skin nerve cells called low-threshold mechanoreceptors. [3] This is a well-studied part of nerve biology. It tells us how touch is felt. It does not by itself show that massage treats any condition.

A second system responds to slow, light stroking. Researchers describe a class of nerve fibers called C-tactile afferents that code for the pleasurable properties of touch. [4] In a small study of 10 people with a rare inherited condition that reduces some thin nerve fibers, gentle slow arm stroking was rated as less pleasant than it was by matched controls. [5] Label: physiological finding, from small studies. It supports the idea that pleasant touch has its own pathway. It does not show a health benefit from massage.

Pain and the Nervous System: Proposed Explanations

A model paper proposed that a mechanical force from manual therapy starts a cascade of responses in the peripheral and central nervous system, and that these then produce the clinical outcomes. [1] Label: proposed model, to guide future studies.

The "gate control" idea, in which touch signals interfere with pain signals in the spinal cord, is often used to explain why rubbing a sore spot helps. A review of localized vibration, which is a device-based stimulus and not hands-on massage, said its pain-relieving action is usually attributed to spinal gate control but is actually more complex. [6] The same review noted that low-frequency, low-intensity stimulation resembling massage may interfere with pain through the limbic system, without a gating mechanism. [6] Label: proposed, with evidence from vibration, not massage.

The Brain

A 2026 scoping review pulled together 47 human studies using EEG, fMRI or fNIRS brain measures. [7] The studies were too different to combine statistically, so the authors summarized them in words, and found that moderate pressure favored a relaxation pattern in autonomic measures. [7] They called for larger, standardized, longer trials to define dose-response, durability and mechanistic mediators. [7] Label: promising but limited; heterogeneous, and measures of brain activity are not the same as health outcomes.

Muscle and Blood

In one 2012 study, 11 young men had massage on one thigh and no treatment on the other after exercise-induced muscle damage; biopsies showed lower levels of some inflammation-related molecules in the massaged muscle. [8] Massage had no effect on muscle metabolites such as glycogen and lactate. [8] Label: preclinical or early laboratory finding; small, young, male, one study. It also does not support the common claim that massage clears lactic acid. See the claim check on toxins and lactic acid.

A 2005 review noted that massage pressure is expected to make muscle more compliant and increase range of motion, and that it might increase blood flow. [2] These are expectations, not demonstrated effects. Proper evidence not found for blood-flow benefit from hands-on massage in the sources we read: we searched PubMed for massage with blood flow and ultrasound measures, and the first results were studies of other treatments (lymphatic drainage in vein disease, an electrical stimulation device and more). For a related claim, see massage and "better circulation".

Stress Hormones

A 2011 review found massage's effect on cortisol was very small and, in most cases, not statistically distinguishable from zero. [9] Its authors concluded that other mechanisms, still to be identified, must account for any clinical benefits. [9] The 2004 meta-analysis found that single sessions reduced state anxiety, blood pressure and heart rate, but not negative mood, immediate pain or cortisol. [10] Label: anxiety effect promising but limited; cortisol explanation not supported.

Putting It Together

Feeling better after a session is real, but pain relief is not proof that tissue has been repaired; see massage and pain science and the longer list of proposed mechanisms.