The Body Keeps the Score: What Is Measured and What Is Myth

"The body keeps the score" is the most quoted sentence about trauma of the past thirty years. It has a precise source: that was the title of Bessel van der Kolk's paper in Harvard Review of Psychiatry in 1994, twenty years before the book of the same name. And it claimed something fairly modest — that physiology changes in post-traumatic stress: arousal, reactions, regulation — not that the body works as an archive of events. The gap between those two readings is the subject of this article.

Where the phrase came from and what it meant

Van der Kolk wrote that traumatic experience is encoded differently from ordinary autobiographical memory: it returns not as a narrative but as a state — sensations, images, readiness to act. Hence the metaphor: a person "remembers" without words. That observation held up. But the metaphor was quickly read literally, and something appeared that the original claim never contained: the idea that a specific event lies inside a specific muscle and can be extracted from it.

What is genuinely known: autonomic tuning

The most solid part here is psychophysiology. The meta-analysis by Pole (2007) pooled 58 resting-baseline studies, 25 startle studies, 17 standardised and 22 idiographic trauma-cue studies. At rest, people with PTSD show higher heart rate (r = 0.18) and skin conductance (r = 0.08); with reminders, heart rate differences are more visible (r = 0.27 in standardised cues), and the most robust markers turned out to be the skin conductance habituation slope and facial electromyography during idiographic cues. The author states the limitation explicitly: the samples skew toward male veterans.

The second layer is heart rate variability. The meta-analysis by Schneider and Schwerdtfeger (2020) found reduced parasympathetic indices at rest (RMSSD and the high-frequency component), a medium effect for overall variability, higher heart rate and a shifted low-to-high frequency ratio. Under stress the same pattern appears, with small effect sizes.

In plain terms: in some people, after prolonged hard experience, the organism stays in a higher state of readiness and returns to rest less easily. That is not tissue memory. It is a setting of regulation — and it moves in both directions. The same setting shows on the emotional side: the wave rises fast and takes a long time to come down — that is unpacked separately in the article on emotional dysregulation.

Bodily reactions to a reminder

The most recognisable experience in this subject is the body reacting before the person has worked out to what. A smell, an intonation, a quality of light, somebody's posture — and the heart is already fast, the palms damp, the stomach cold. That is exactly the effect trauma-cue studies measure: the reaction starts before the explanation.

Two practical consequences follow. First, the sensation is not proof that "a memory came out of the body"; it is an alarm firing on a coarse match of features. Second, arguing with it does not work, because it starts below the level of reasoning — as in the fight, flight and freeze responses. What the extreme form of such a return looks like is covered in the article on flashbacks; and when the response goes the other way, into shutdown and watching yourself from outside, that is dissociation.

Sleep, tension and what gets called "holding patterns"

Sleep is the best-measured bodily line. The meta-analysis by Kobayashi et al. (2007) across 20 polysomnographic studies found more light stage 1 sleep, less slow-wave sleep and greater rapid-eye-movement density in people with PTSD. So the complaint "I sleep but do not rest" has an objective correlate, not only a subjective one.

Muscle tension is a more honest and duller story. Persistent tension in the shoulders, jaw and diaphragm is common in people with high background arousal, and it makes sense: tension is part of readiness to act. But "maps" assigning an event or a feeling to each muscle have never been confirmed in research. The body can be measured; a biography cannot be read out of it.

Where retellings run past the data

Four claims frequently attributed to the book that the research does not support:

• "Trauma is stored in tissue and can be released from it." Memory of an event is a brain function. What is measured in the body is a response, not content.
• "The body remembers exactly, like a recording." The opposite: the bodily response fires on a coarse match of features, which is why it often misfires and reacts to what is safe.
• "Work through the body and the trauma will come out." No method promises release as an event. Improvement looks like a lower background and more manageability, not a one-off liberation.
• "Illness is unprocessed trauma." Associations between hard childhood experience and physical health are real, but they are risk statistics in large groups, not an explanation of one person's diagnosis.

What this means for somatic trauma work

Body-oriented approaches are neither quackery nor a proven standard; they sit in between, and naming that honestly matters more than picking a side. The review by Kuhfuß et al. (2021) on somatic experiencing selected 16 papers out of 83 and described the result as preliminary evidence of positive effects on symptoms, with mixed study quality and a need for unbiased randomised trials.

For comparison, trauma-focused psychotherapy has an incomparably thicker base — see the article on EMDR and the general account of what PTSD is. A reasonable position: body work as support for regulation and a way back into contact with the present, yes; body work as a replacement for trauma therapy, no.

One request to ourselves as well: go carefully with promises of growth. Why "trauma made me stronger" is a harmful formula is unpacked separately in the article on post-traumatic growth.

When it is not trauma

A high bodily background is non-specific. Before explaining it by the past, it is reasonable to walk through the neighbours.

• Depression. Produces heaviness, exhaustion and pain in the body, but is usually episodic and lacks a traumatic core — no intrusions, no avoidance of reminders, no sense of current threat.
• Long-running exhaustion. Years of overload reproduce almost the whole bodily half of the picture and respond to unloading rather than to work on the past.
• Anxiety disorder. Palpitations, a lump in the throat, tension are its ordinary set; the difference is that the worry is not tied to reminders.
• A borderline pattern. Sharp bodily waves occur here too, but the leading theme is instability of self-image and fear of abandonment.
• ADHD and autism. Motor restlessness and sensory overload run evenly from childhood and are explained by how attention and perception are built, not by events.
• Bodily causes. A shifted TSH produces palpitations, tremor and sweating; low ferritin produces restlessness at rest and flatness; lack of vitamin B12 produces numbness, tremor and brain fog. Add apnoea, sleep debt, caffeine and alcohol withdrawal. Results you already have, the service will help you read — and then it is clear what the body explains and what it does not.

What to actually do with the body

The working frame is not "release the past" but "come back into the present and lower the background".

• Grounding. Footing, five objects named aloud, cold water, a change of posture. The point is to mark that now is not then and let the wave pass.
• Breathing with a long exhale. A simple way to influence the very parasympathetic side that heart rate variability measures. Not treatment — a one-minute tool.
• Regular movement. Walking, swimming, resistance work: predictable load lowers background arousal better than talking about it.
• Sleep. Stable hours, darkness, no screen in bed. Given the polysomnography data this is not hygiene for its own sake — it is the one line where improvement is felt quickly.
• Boundaries around contact. Less of what reliably knocks you over, and an exit from the situation agreed in advance.

If you want to see where all of this adds up, there are two different instruments. The free PCL-5 scale is a screener for PTSD symptoms: one axis, one score. The paid complex PTSD test separates the traumatic core from disturbances in self-organisation across six clusters and looks separately at the cost in work, relationships and daily life. Neither is a diagnosis — a clinician makes that in person.

If things are hard right now — thoughts of not living, a pull to hurt yourself, or ongoing danger — in the US call or text 988 (Suicide and Crisis Lifeline); in the UK, Samaritans on 116 123; elsewhere, your local emergency number.

This article is informational and does not diagnose. It does not claim that you have PTSD, nor that you do not: that is determined by a clinician in person.