A recent theoretical paper published in Frontiers in Systems Neuroscience suggests that psychological trauma is not literally stored in the tissues of the body. Instead, the authors propose that trauma creates a rigid pattern of threat prediction within the brain, reducing cognitive flexibility. This updated perspective provides evidence that therapies focusing on mental state shifting, such as flow states, may help retrain the nervous system and support recovery.

In 2014, psychiatrist Bessel van der Kolk published The Body Keeps the Score, a book suggesting that trauma alters the nervous system and becomes physically lodged in the body. The bestseller popularized the idea that individuals cannot simply talk through trauma, as the body continues to react to past threats. While the book gained immense public and clinical popularity, some scientists have criticized its underlying biological mechanisms.

Scientists Steven Kotler, Michael Mannino, Glenn Fox, and Karl Friston recently authored a paper to address these mechanistic discrepancies. They argue that the popular metaphor of somatic storage is biologically inaccurate. They propose an alternative model based on computational neuroscience.

Michael Mannino, the chief science officer of the Flow Research Collective and a distinguished research fellow at Florida Atlantic University, explained the motivation behind the paper. “Two things motivated the pushback,” Mannino said. He noted a conversation with researcher George Bonanno regarding resilience data.

“Bonanno had looked at some of the same kinds of trauma data and drawn very different conclusions than the dominant ‘trauma is stored in the body’ framing,” Mannino explained.

This conversation resonated with his own observations. “That conversation crystallized something I had already been sensing: the model did not line up with what we were seeing in flow research or in real-world performance contexts,” Mannino said.

Additionally, the authors noticed conflicts between the body storage theory and their own work on optimal mental states. “Second, the ‘body keeps the score’ framing became harder to reconcile with the evidence around flow,” Mannino noted. “Flow is common, trainable, and repeatedly associated with improvements in psychological functioning.”

He elaborated on this contradiction. “If trauma were literally stored in somatic tissue in the way the popular metaphor suggests, then it would be strange that flow training could have such broad effects on trauma recovery,” Mannino said. “Flow is not a specialized somatic therapy. It is not manually targeting hidden trauma deposits in muscle or fascia.”

“So if flow is helping people shift traumatic patterns, then the problem likely is not ‘stored trauma’ in the body,” Mannino continued. “It is more likely a problem in how the brain predicts threat, safety, agency, and action.”

The authors also felt the cultural timing was right for a shift in perspective. “The timing also matters because the popular culture around trauma has shifted,” Mannino pointed out. “Many people now lead with trauma as identity. The concern is that the storage metaphor may unintentionally reinforce victim identity, external locus of control, and chronic reactivation of painful memories.”

“If trauma is framed as something buried inside the body that must be located and released, people can feel less empowered,” Mannino added. “But if trauma is understood as a disorder of prediction, then we have more actionable targets.”

In their paper, the authors reframe trauma as a disorder of prediction rather than a disorder of storage. They use a concept called predictive coding, which suggests the brain constantly guesses what will happen next based on past experiences. In a healthy brain, these predictions are flexible and update when new information arrives.

Following a traumatic event, the brain tends to assign too much weight to signs of danger. This creates a highly rigid system where the brain anticipates threats everywhere, leading to hypervigilance and avoidance. The brain misinterprets regular physical sensations, like a racing heart, as proof of immediate danger.

Because of this rigid prediction system, the brain loses what scientists call metastability. Metastability refers to the brain’s ability to fluidly switch between different networks and mental states. High metastability allows for cognitive flexibility, meaning a person can easily adapt to new situations.

Trauma traps the brain in a narrow, inflexible state of fear. The body acts as a messenger in this process, sending signals that the brain misinterprets, but the body does not serve as an archive for the trauma itself.

“The main takeaway is empowering: trauma is not necessarily something hidden in the tissues that must be excavated,” Mannino said. “It may be better understood as a maladaptive prediction system.”

He explained that the brain is always forecasting safety, danger, and the meaning of physical sensations. “The brain is constantly predicting what is safe, what is dangerous, what matters, what action is possible, and what the body’s sensations mean,” he said.

“In trauma, those predictions can become rigid, overgeneralized, and threat-biased,” Mannino noted. “The person may not simply be ‘remembering’ the past. Their nervous system may be predicting the present and future through the lens of unresolved danger.”

This distinction is important for treatment. “That distinction matters because storage is a very hard target,” Mannino said. “We do not fully understand how memories are stored in the brain, let alone how traumatic memory would be stored in the body in a literal biological sense. But prediction is a much more tractable target.”

Mannino emphasized that psychology already utilizes tools that address prediction, including cognitive reframing, exposure therapies, attention training, and mindset changes. “So the average person should not hear this as ‘the body does not matter,’” he said. “The body absolutely matters. Pain, posture, breath, autonomic tone, movement, touch, and interoception all shape prediction.”

“But the mechanism may not be that trauma is physically stored in the muscles,” Mannino continued. “The mechanism may be that bodily signals are feeding into the brain’s predictive machinery, shaping what the person expects, fears, avoids, or interprets as dangerous.”

To help restore mental flexibility, the authors suggest flow states could act as a powerful intervention. Flow happens when a person becomes completely absorbed in a meaningful, highly challenging activity. Action sports, playing music, or engaging in complex tasks can trigger these states.

During flow, the brain’s threat detection centers quiet down, and networks related to focus and adaptation engage. “The strongest evidence begins with a large body of converging observations,” Mannino said regarding the connection between flow and trauma recovery.

He pointed to action sports athletes who face high risk but often report feeling regulated and empowered. “If trauma were simply stored in the body through overwhelming experience, then repeated exposure to danger should reliably worsen trauma,” Mannino noted. “But that is not always what we see.”

“Another line of evidence comes from flow-based and adventure-based interventions,” Mannino explained. Programs focusing on outdoor leadership or addiction recovery tend to help individuals break out of fixed behavioral loops. “Flow may help because it trains flexible state-shifting, agency, attention, and adaptive prediction under challenge,” he said.

“The strongest theoretical argument is that flow targets the prediction system,” Mannino added. “Flow changes attention, motivation, agency, threat perception, self-processing, and action-readiness. Those are all central to trauma.”

“If trauma involves rigid threat prediction, then flow may help by creating conditions where the nervous system learns, experientially, that challenge can be navigated safely and effectively,” Mannino said.

While the authors challenge the idea of somatic storage, they do not suggest that body-based therapies are ineffective. Bodywork, massage, and breathing exercises often provide significant relief for trauma survivors. The authors simply propose a different biological explanation for why these treatments work.

“We should not dismiss the therapeutic benefit people report,” Mannino stated. “Bodywork, somatic therapies, massage, acupuncture, breathwork, and related interventions clearly help many people. The question is not whether they can help. The question is why they help.”

Mannino suggests that finding a tight or painful spot during bodywork might create a prediction error in the brain. “A bodyworker finds a spot of pain, tightness, or unusual sensation,” he explained. “The brain then has to interpret that signal: ‘What is causing this?’”

Because sensation and emotion are linked, the brain might generate a memory or narrative to explain the physical feeling. “That does not mean the memory was literally stored in that muscle,” Mannino said. “It may mean the sensation created a prediction error, and the brain searched for a prior, an explanatory model, or an associated emotional memory.”

He believes this perspective opens the door to better scientific questions. “Is the therapeutic effect coming from touch? From relaxation? From parasympathetic activation? From interpersonal synchrony? From human connection? From increased interoceptive awareness? From changing threat predictions? From the therapist-client relational field?” Mannino asked. “The goal should not be to banish bodywork from trauma recovery. The goal should be to get the mechanism right so we can improve the interventions.”

The authors acknowledge limitations in their current model. Their framework represents a proposed mechanistic reframing, not a finalized clinical doctrine. Additionally, the specific connection between flow states and trauma recovery requires further direct testing.

“What remains speculative is the exact mechanism,” Mannino noted. “We do not yet have definitive evidence that flow heals trauma by increasing metastability, altering prediction error, or reorganizing attractor dynamics. Those are plausible, testable hypotheses, not settled facts.”

Trauma also comes in many forms, meaning this model might not apply universally. “PTSD, developmental trauma, acute trauma, grief, moral injury, chronic stress, and traumatic brain injury may involve overlapping but distinct mechanisms,” Mannino cautioned. “A prediction-based model may be powerful, but it should not be treated as a one-size-fits-all explanation.”

“A second caveat is that body-based therapies may still work,” Mannino added. “Rejecting the storage metaphor does not mean rejecting massage, somatic therapy, breathwork, movement, or bodywork. It means we should ask better mechanistic questions. These practices may work by changing autonomic state, increasing interoceptive precision, reducing threat prediction, enhancing safety cues, promoting interpersonal synchrony, or creating prediction error that allows updating.”

The evidence connecting flow to healing also needs formal neuroscientific backing. “Finally, the flow-trauma connection is still developing,” Mannino noted. “There is strong theoretical and anecdotal support, plus related evidence from performance, adventure therapy, and clinical psychology. But the direct neuroscientific evidence still needs to be built.”

Moving forward, the researchers hope to test their claims by measuring brain dynamics in populations with trauma histories. “To test the claim directly, we would need to measure brain dynamics in trauma populations, especially looking at whether PTSD is associated with reduced metastability, reduced flexibility, or difficulty transitioning between neural states,” Mannino said.

A strong study might use brain imaging to look at flexibility in people with trauma compared to a control group. “The prediction would be that PTSD involves overly rigid attractor dynamics: the nervous system gets stuck in certain threat-biased patterns and has trouble transitioning flexibly into alternative states,” Mannino explained.

Researchers could also test different interventions, comparing body-based treatments to flow-based activities. Mannino mentioned a speculative idea from Kotler involving bodywork. “If a bodyworker stimulates a painful area and the person reports a memory or emotional response, researchers could examine whether the content of that response varies according to contralateral brain-body organization,” Mannino said.

“So yes, the line of research is moving toward empirical testing,” Mannino emphasized. “The key is to move beyond metaphor and into measurable predictions: metastability, complexity, state transitions, prediction error, symptom change, and intervention response.”

The broader goal is to build a performance-based approach to neuroscience. “The next step is to test the model directly,” Mannino stated. “One direction is to examine metastability and neural complexity in trauma populations, especially before and after interventions.”

“A second direction is to compare somatic-based interventions with flow-based interventions, and potentially with combined interventions,” Mannino continued. “If bodywork helps, we want to know why.”

“A third direction is to study prediction-related mechanisms in adjacent conditions: PTSD, depression, anxiety, addiction, traumatic brain injury, Parkinson’s, ALS, and other neurodegenerative or neuropsychiatric conditions where rigidity, loss of agency, and impaired state transitions may play a role,” Mannino added.

The team envisions a future where treatments focus on expanding mental capabilities. “The broader aim is to develop a performance-neuroscience approach to brain health: not just reducing symptoms, but restoring flexibility, agency, adaptive prediction, and the capacity to enter high-functioning states like flow,” Mannino concluded. “That is where this line of research is headed.”

Scheduled for release in July, Kotler and Mannino’s forthcoming textbook defines the emerging scientific field of performance neuroscience. The book explores how the brain and body coordinate under high-pressure conditions, explaining the biological mechanisms behind flow states, stress regulation, and peak human achievement. The book will “will help formalize the field and provide a broader scientific foundation for this line of work.”

The study, “The body does not keep the score: trauma, predictive coding, and the restoration of metastability,” was authored by Steven Kotler, Michael Mannino, Glenn Fox, and Karl Friston.