The Science of Somatic Regulation: What Clinicians Need to Know About Polyvagal Theory Today
Understanding the new conversation around Poly vagal Theory
Polyvagal Theory has shaped trauma therapy for decades, but recent critiques have created confusion about what is accurate, what is debated and what still holds true. This blog is my attempt to understand and offer a clear, research informed breakdown of the critiques, Stephen Porges’ responses, and the emerging neuroscience of interoception and predictive processing.
Polyvagal Theory has shaped an entire generation of trauma therapists. For many of us, it gave language to something we already sensed in the room, namely, that safety, connection, and autonomic state are deeply intertwined. But like any influential model, it has also drawn increasing critique from academic neuroscience, psychophysiology and clinical research. These critiques don’t dismiss the role of the body, instead, they challenge several of Polyvagal Theory’s foundational claims including:
The idea of a neat evolutionary hierarchy within the vagus
Critics such as Grossman & Taylor (2024) argue that PVT overstates the phylogenetic separation between dorsal and ventral vagal pathways and misinterprets comparative anatomy.Polyvagal Theory’s broad interpretation of dorsal vagal “shutdown”
Researchers have noted that immobilisation behaviours are not solely vagally mediated, and that dorsal vagal activation does not necessarily equate to collapse, shutdown or fainting (Zeisler & Brown, 2023).The assumption that heart-rate variability (HRV) is a clean, direct index of “vagal tone”
HRV researchers consistently point out that HRV is multiply determined by breathing rate, sympathetic activity and mechanical factors—and cannot be used as a singular biomarker for vagal functioning.
Importantly, Stephen Porges has publicly responded to each of these critiques. He clarifies that Polyvagal Theory was never intended as a biomarker based diagnostic tool. Rather, it is a theoretical model for understanding how neuroception and autonomic shifts shape behaviour. In recent publications and interviews (Porges, 2022–2024), he has emphasised that many critiques conflate PVT with “polyvagal inspired practices” or misapply the theory beyond its intended scope “Polyvagal Theory is a conceptual model, not a measurement system” (Porges, 2021, Frontiers in Psychiatry).
In interviews, he consistently emphasises that many critiques arise when people treat PVT as a protocol or physiological test rather than a framework. As he put it, “The theory is often misunderstood or misapplied when people want it to be something it was never intended to be” (Porges, 2022 interview, Therapy Matters).
He also explicitly acknowledges the limitations of HRV and that HRV is not a direct readout of vagal states but a proxy influenced by multiple interacting variables. stating, “Heart rate variability is not a measure of vagal tone. It is influenced by many sources” (Porges, 2019, Psychophysiology). His view is clear and the theory remains intact, but it must be used for what it is rather than what we might wish it to be.
At the same time, other fields have rapidly expanded. Research on interoception has sharpened our understanding of how sensations become feelings, and predictive processing has reframed emotion as the brain’s ongoing attempt to anticipate the body. These perspectives often emphasise top down prediction more strongly than bottom up sensory input, which leaves many clinicians wondering whether somatic therapies are still scientifically justified or whether the body’s influence is being minimised or lost.
A closer look at the evidence tells us a different story. Even where Polyvagal Theory is being revised or questioned, the core principle it helped popularise remains robust: sensory input from the body profoundly shapes regulation. All day long, our body feeds information upward through mechanoreceptors, interoceptive fibres, baroreceptors and proprioceptive pathways. These quiet internal signals shape how we feel, how safe we sense ourselves to be and how stable our physiology stays. Predictive models do not cancel out the body’s role. They simply show us that regulation is a conversation: the body sends signals up, and the brain responds with expectations shaped by history, conditioning attachment, trauma and memory.
So what’s really shifting? It’s not the recognition that the body matters, but there is a more nuanced understanding of how it matters and how bodily input interacts with cognitive, relational and developmental forces. With the science we have today, I believe somatic therapies are not only relevant… they are needed. Their effectiveness rests on the way they may alter the quality, intensity and rhythm of interoceptive signals the brain must account for, creating opportunities for the brain’s outdated, threat based predictions to be revised.
Rather than choosing between Polyvagal Theory and predictive processing, or body over brain, it’s more accurate to see regulation as a dynamic body to brain loop, where sensation and interpretation shape one another continuously. Very simply this means, we work with both the body and the brain because we need both!
This blog aims to explores that integration. What Polyvagal Theory contributes and gets right, where the critiques are valid, and how contemporary research deepens our understanding of regulation in trauma therapy today. If anything perhaps this new research invites us toward a richer realisation…That regulation is an ongoing conversation between the body and the brain. Sensory input, prediction, visceral feedback, relational cues and meaning all co create our moment to moment experience of safety or threat.
What Polyvagal Theory Gets Right and What Has Been Critiqued
Polyvagal Theory (PVT) has given many of us something we had been searching for… a coherent language for the things we were feeling in the room long before we had the neuroscience to back it up. It has helped us describe how the body responds to threat before we can think, how connection can stabilise our physiology, and how different autonomic patterns can be mapped in simple states safety, mobilisation or collapse. These insights continue to resonate because they reflect what people experience and what clinicians witness every day.
At the same time, Polyvagal Theory has been scrutinised in academic circles, not because the body doesn’t matter, but because researchers want to refine the mechanisms that explain how regulation truly works. I believe these debates actually deepen (not weaken) the case for somatic therapy.
PVT was right about the vagus being mostly sensory (afferent)
One of PVT’s most important contributions was emphasising that the vagus nerve carries far more information upward to the brain than downward to the body. Around 80% of vagal fibres are afferent (Agostoni et al., 1957; Breit et al., 2018). This bottom-up flow of information is widely accepted across physiology and remains essential to understanding interoception, emotional experience and autonomic regulation.
Seth’s (2012) formulation captures this beautifully: “Interoceptive inference conceives of subjective feeling states as arising from actively inferred generative (predictive) models of the causes of interoceptive afferents.” Translation: Interoceptive afferents are the signals coming from inside the body, things like heartbeat, breath, gut sensations and muscle tension. They travel upward through nerves like the vagus to the brain.
Predictive models of the causes means that the brain has internal expectations or “best guesses” about:
What the body is doing
Why sensations are happening
What they mean
These predictions are shaped by past experience which could include, trauma, learning or attachment experiences as a few examples.
So the brain tries to guess what the body’s signals mean and this means as subjective feeling states that arise form these predictions, we feel things like fear, excitement, anxiety or states of calm that are shaped by the brains brain’s interpretation of the body’s signals.
Same body signal → different brain prediction → different feeling. For Example:
Fast heartbeat + “danger prediction” → panic
Fast heartbeat + “exercise prediction” → energy/excitement
Emotion = body’s signals + brain’s interpretation.
If we put this all together this means that your emotions don’t just come only from your body… but they also don’t just come only from your brain. They come from the brain predicting what the body’s signals mean. The body sends information up, the brain predicts and interprets down. Feeling states emerge from their interaction.
PVT was right that relational cues regulate physiology
Even the strongest critics of PVT agree that human connection shapes our autonomic state. Warm tone of voice, facial expression, posture, orientation, proximity, and co-regulation have measurable effects on heart rate, HRV, parasympathetic tone and emotional stability. Decades of attachment research, developmental science and affective neuroscience support this.
“Co-regulation refers to a process in which partners’ emotions are bi-directionally linked and mutually dampening … ultimately contributing to emotional stability.” (Reed & Ebberwein, 2015, p. 2)
“Warm gaze, gentle eye contact, warm reassuring voice, and open body posture communicate safety … Regulation can occur through self-regulation or co-regulation.” (Žvelc & Žvelc, 2025, p. 5)
Supporting studies:
Golland et al. (2015) discusses Autonomic synchrony through mere co-presence
Bornstein (2023) looks at multilevel co-regulation via facial expression, voice, proximity
Hietanen et al. (2019) shows how gaze and facial cues change autonomic output
Porges (2022) Look starting how social engagement system modulates vagal state
Shutdown and Collapse: Clinically Useful Language, Evolutionarily Debated
Many therapists appreciate PVT’s distinction between fight/flight and collapse/shutdown because it mirrors many peoples lived experience. However, the evolutionary hierarchy proposed by Porges eg. The idea that mammals have a uniquely evolved “ventral vagal system” that enables prosocial calm is the area most often challenged.
What the critique actually says
Comparative anatomy shows that many non-mammalian species (fish, reptiles, amphibians, birds) demonstrate:
Respiratory sinus arrhythmia (RSA)
Vagal modulation of heart rate
Coordinated cardio respiratory rhythms
These animals do not possess the mammalian “ventral vagal complex,” yet they show similar patterns of parasympathetic regulation.
The debate is that:
Cardio-respiratory coupling is not unique to mammals.
It cannot be used as evidence for a uniquely specialised social engagement vagal circuit.
The evolutionary story is more complex than PVT proposes.
This critique doesn’t dismiss vagal regulation, but it does challenge the phylogenetic framing. The underlying physiology remains valid.
HRV is useful but not a pure measure of vagal tone
PVT sometimes treats heart rate variability as a direct index of ventral vagal function. Critics rightly point out this is not accurate.
HRV reflects:
Respiration
Baroreflexes
Sympathetic activity
Parasympathetic activity
Metabolic influences
(Billman, 2013)
So yes, HRV still tells us something about regulation but it cannot be treated as the one and only clean biomarker of vagal tone.
Shutdown and dissociation are multi network, whole body events, not single-pathway vagal reflexes
This is one of the most important scientific updates. Shutdown does not arise simply from “dorsal vagal activation.” Studies by Ruth Lanius and colleagues show that dissociation involves network disconnection across different brain regions that include:
Insula
mPFC
ACC
PAG
These networks regulate body awareness, emotion, movement and threat appraisal. When they fall out of sync, people can experience numbness, disconnection, or a feeling of “going away.” Freeze is also more complex than it was originally framed. It often involves high sympathetic arousal alongside immobility. This is sometimes described as having the “accelerator and brake’ ( sympathetic activation along with dorsal activation) on together at the same time. This helps explain why people who have experience trauma may feel fear internally while appearing calm externally.
In practice, shutdown and dissociation are not single pathway reflexes. They involve the whole brain and the whole body, which is why clients often describe them as total states of being rather than isolated reactions They involve changes in brain networks that regulate awareness, emotion, movement and threat detection. This more detailed understanding helps us see why somatic therapies need to work with bodily sensation, orientation, movement and relational safety not only with ‘vagal tone’ or the vagus to support recovery from trauma.
Neuroception is clinically useful… but scientifically a concept, not a mechanism
Porges’ term neuroception beautifully captures the idea of automatic, unconscious detection of safety or danger. Clinically, it resonates. Scientifically, critics emphasise:
There is no biomarker for neuroception.
No scan or test can identify it.
Nothing distinguishes it from known mechanisms like amygdala activity, insula processing, or brainstem threat circuits.
Safety/danger detection is distributed across many networks and these networks vary depending on context, trauma history, and sensory input. This means that there is no single “safety detection system” and we can think of nueroception as a higher level concept vs a defined circuit.
“There is no single fear circuit. Defensive responses emerge from multiple interacting neural systems depending on context and prior learning.”
(LeDoux, 2012)
“Trauma-related responses reflect disruptions across interconnected networks …rather than a single threat-processing circuit.”
(Lanius et al., multiple publications)
“Safety signalling engages multiple cortical and subcortical structures … represented by a distributed neural architecture.”
(Christianson et al., 2012)
“Emotional and threat-related responses arise from distributed hierarchical systems integrating interoceptive predictions and sensory evidence.” (Seth & Friston, 2016)
Together, these findings support a simple, modern conclusion: Neuroception is a helpful umbrella term, not a standalone biological system. Researchers note that the core idea behind neuroception being a pre-conscious evaluation of safety or danger is already recognised in neuroscience so they argue that calling these processes “neuroception” doesn’t add something new scientifically, even though the term is helpful clinically. Some critics also warn that the way neuroception is used in clinical work can suggest there is a single, dedicated mechanism for detecting safety or danger however research shows the opposite.
What this all means for the field
Taken together, these critiques don’t undermine Polyvagal Theory’s clinical usefulness, they enrich it.
Polyvagal Theory captured essentials:
The body’s primacy
The social nature of regulation
The reality of shutdown and collapse
The role of autonomic state in emotion
What is shifting is our understanding of how these processes work on a neural level. Contemporary neuroscience points toward a more integrated mode, one that includes:
Interoception
Predictive processing
Brainstem circuitry
Distributed networks
Sensory pathways
Attachment and relational safety
Developmental learning
It’s clear that regulation emerges from moment to moment interactions across many systems, not from a single pathway. The central truth remains unchanged, and strongly supported, regulation is deeply body based.
How Regulation Really Works: The Body–Brain Loop
One of the most important shifts in trauma neuroscience is the recognition that regulation doesn’t come only from the brain or only from the body. It comes from a continuous moment to moment conversation between the two. Early Polyvagal Theory helped clinicians pay attention to autonomic state and vagal pathways. More recent research in interoception, predictive processing and affective neuroscience has expanded that picture, showing that regulation actually emerges from the interaction of top-down predictions and bottom-up sensory signals. Neither direction is sufficient on its own.
Hers a look at how this loop works
Bottom up sensory input from the body shapes the brain’s state
The body is constantly talking to the brain. Mechanoreceptors, proprioceptors, baroreceptors, interoceptive C-fibres and vagal afferents send a steady stream of signals upward into the nucleus tractus solitarius (NTS), parabrachial nucleus, insula, anterior cingulate cortex (ACC) and other brainstem centres involved in arousal, emotion and threat appraisal. Studies by Craig (2002), McGlone et al. (2014), Garfinkel et al. (2014) and Breit et al. (2018) all show that touch, breath, heart to lung coupling and visceral sensation directly modulate autonomic function and emotional experience. In other words, the body is not passive. It continuously shapes, constrains and informs the brain’s sense of what is happening and what is safe.
Top down means the brain predicts and interprets bodily signals
At the same time, predictive processing models (Feldman Barrett, Friston, Seth) show that the brain is not simply waiting for sensation to arrive. It is constantly generating anticipatory predictions about what the body is likely to feel. These predictions colour our experience of sensation turning the same flutter in the chest into excitement for one person and panic for another.
But crucially these predictions are not fixed stories. They are hypotheses. The brain revises them when new sensory evidence contradicts what it expected. This is the heart of therapeutic change. When we can learn to takes a slower breath, ground through our feet, or finds rhythm and movement in the body, we generate sensory signals that do not fit the brain’s old threat based expectations. This creates a prediction error, which a fancy scientific name for what we call learning.
Regulation emerges from the negotiation between prediction and sensation
Neither the body nor the brain “wins.” Regulation arises through their ongoing dialogue:
The brain predicts based on history and survival learning.
The body sends data based on current physiological state.
When the two don’t match, the system adjusts, shifting arousal, emotion and meaning. If the new sensory cues communicate safety, the brain gradually reorganises itself around them.This explains why someone can feel unsafe in a perfectly safe environment: the prediction of danger is stronger than the incoming evidence. It also explains why somatic interventions work so reliably. When we change the sensory input, we change what the brain has to work with.
Trauma disrupts the loop
Trauma impacts both sides of this regulatory conversation.
Bottom up disruptions mean:
Chronic sympathetic activation
Blunted or fragmented interoception
Collapse or shutdown
Top down disruptions mean:
The brain begins to expect danger
Interpretations become rigid
Benign sensations feel threatening
Together, this creates a narrow window of tolerance and a heightened sensitivity to cues that might signal threat.
Somatic therapy can help restore flexibility in the loop
Somatic therapy intervenes directly in the conversation between body and brain. By shifting breath, posture, movement, orientation, touch and relational cues, we can change the sensory input rising from the body. When the body sends different information for example: slower, steadier, more rhythmic sensations, the brain isnt going to continue predict danger in the same way. Threat models can soften and something new is possible.
Co-regulation also amplifies this. An attuned voice, warm presence, soft gaze or grounded posture offers the nervous system relational evidence of safety, something the brain takes seriously.This is why somatic interventions are not optional add on’s. They are a direct way of working with trauma because they engage the neural systems that generate emotional experience and autonomic state. They give the brain the opportunity to update the story it has been telling for years: that danger is always near.
“Coregulation involves bidirectional linkage of physiological and behavioural processes between partners and reflects the ways one person’s nervous system helps to regulate another’s.” (Bornstein, 2023)
Just being together (presence, gaze, posture) can even align our physiology.
“We found that the autonomic signals of co-present participants were idiosyncratically synchronized and that the degree of this synchronization was correlated with the convergence of their emotional responses.” (Golland et al., 2015)
See full article: Golland, Y.; Levit-Binnun, N. et al. (2015). “The Mere Co-Presence: Synchronization of Autonomic Signals …” PLoS ONE, 10(5): e0125804.
Voice tone can also be a nervous system cue
“Caregiver vocalisations drive fluctuations in infant autonomic arousal, demonstrating that prosodic voice modulates the listener’s nervous system.” (Wass et al., 2022)
How This Helps Us Understand Shutdown, Freeze and Dissociation More Accurately
For a long time we relied on simple categories to describe trauma responses: Fight, flight, freeze, shutdown, as if each state were driven by a single switch inside the autonomic nervous system. Polyvagal Theory helped expand that view by naming immobilisation as a legitimate survival response, not a failure of will. But newer neuroscience shows these states are not caused by one vagal pathway alone. Instead, they arise from multiple, interacting neural networks that come online in different configurations depending on context and history.
This gives us a far more accurate and compassionate map of what our clients are experiencing when they freeze, collapse or dissociate.
Shutdown and dissociation involve whole brain network changes, not just vagal dominance
Shutdown does not happen solely because the dorsal vagal complex switches on. Trauma-related neuroimaging consistently shows that dissociation reflects network-wide disconnection across regions responsible for awareness, emotion and self-connection, especially the insula, medial prefrontal cortex (mPFC), and periaqueductal gray (PAG) (Lanius et al.).
When these areas fall out of synchrony under overwhelming threat, clients feel numb, disconnected or “far away,” even if the body looks calm. In simple terms:Dissociation isn’t a single vagal reflex. It’s a whole-brain protective strategy.
Freeze is not “parasympathetic only”, it is sympathetic arousal + immobility
Freeze is often misunderstood as a purely parasympathetic collapse. In fact, research by Schauer & Elbert (2010) and others shows that freeze typically involves:
High sympathetic activation internally,
Combined with outward immobilisation.
This is why clients in freeze often say things like “I couldn’t move” while simultaneously reporting feeling tense, hyperaware, or terrified. In simple terms:Freeze is the accelerator and the brake pressed at the same time.
Shutdown is shaped by both top-down predictions and bottom-up overload
Shutdown happens when the brain predicts that:
Escape is impossible
Fight or flight will fail
Immobilisation is the safest option
As discussed, these predictions come from developmental learning, attachment history and prior trauma. Importantly, shutdown also emerges from bottom-up survival signals like intense heart rate surges, breath restriction, interoceptive overload, proprioceptive overwhelm. When the body signals “too much,” the brain can shut down awareness to protect the system. Agin, this means shutdown is created through the interaction of: both top down meaning and prediction and bottom up sensory overwhelm.
Why somatic therapy is IMPORTANT for working with freeze, shutdown and dissociation
Understanding these states as multi network, multi layered processes explains why somatic therapy is important, especially when cognitive based work simply cannot reach these places.
Somatic therapy can help by:
Increasing interoceptive accuracy through gentle tracking
Restoring insula to mPFC connectivity with slow breath and grounding
Shifting sympathetic to parasympathetic balance through posture, movement and orientation
Offering co-regulation to reduce limbic threat signalling
Enabling small, safe motor impulses to counter immobilisation
Helping clients renegotiate freeze without overwhelming the system
When someone is dissociated or frozen, words are not enough The body becomes the only doorway back to connection and regulation.
A more accurate, research-grounded map of shutdown
Blending neuroscience with therapy insights, shutdown and dissociation can be understood as states where:
Sensory input overwhelms the system
Predictions of danger remain rigid and unchallenged
Multiple networks disconnect to reduce suffering
Autonomic responses shift into extreme protection
Movement and perception narrow to conserve energy
This is not pathology. It is intelligent survival. With this understanding, clinicians can support people in therapy by safely pacing sessions, supporting titration, and using somatic resources to gradually restore connection, movement and choice. Shutdown is not an enemy to fight. It is a system doing its best to keep someone alive.
Making Sense of Sensation vs Prediction: Does the Body Come First or the Brain?
One of the most active areas of debate in contemporary neuroscience concerns what comes first, is it bodily sensation or the brain’s interpretation. Predictive processing models argue that the brain continually generates predictions about the body, shaping how our sensations are perceived. Somatic therapies, on the other hand, emphasise the role of bottom-up sensory input in shifting physiological state. Research now shows that the answer is not either or, but that regulation emerges from a reciprocal loop between prediction and sensation.
The brain predicts first, but only as a hypothesis
Predictive coding research shows that the brain continuously forms anticipatory models of what the body will feel, based on past experience, attachment history and survival learning (Feldman Barrett; Friston; Seth). These predictions occur faster than conscious awareness and influence how incoming signals are interpreted. However, predictions are not final or authoritative. They function as the brains best guesses, which must be confirmed or corrected by sensory evidence. In simple terms: The brain starts the story, but the body finishes it.
Sensory receptors continuously send data that update or override predictions
While the brain generates predictions, sensory receptors across the body convert our moment to moment experiences and changes into electrical signals that rise to the brainstem and higher regulatory centres. This process called mechanotransduction is well established in physiology (Delmas & Coste, 2013; Abraira & Ginty, 2013).
For example:
Baroreceptors detect blood pressure changes and send signals that adjust heart rate within milliseconds.
Mechanoreceptors and proprioceptors inform the brain about posture, movement and muscle tone, shaping arousal and stability.
Interoceptive C-fibres and vagal afferents communicate the state of the gut, heart, lungs and viscera to the insula and ACC.
These pathways provide the brain with “real-time data” that can contradict, refine or reinforce its predictions.
Sensory input does influence regulation, this is a scientific fact
There is robust evidence that bottom-up sensory input alters autonomic state:
Slow exhalation increases baroreceptor firing and reduces sympathetic arousal (Garfinkel et al., 2014).
C-Tactile afferent touch reduces amygdala activity and increases parasympathetic tone (McGlone et al., 2014).
Proprioceptive stimulation affects brainstem arousal circuits and stabilises autonomic output (Proske & Gandevia, 2012).
Vagal afferent stimulation alters inflammation, mood and autonomic reactivity (Breit et al., 2018).
These are foundational physiological facts, not theoretical interpretations. The body continuously shapes the brain’s regulatory state.
Regulation comes from the negotiation between the two. The most accurate scientific model is this:
The brain predicts what the body will feel.
The body sends back data that confirm or challenge the prediction.
The difference (prediction error) drives adjustments in autonomic state, emotion and perception.
New sensory experiences, especially safe bodily experiences can gradually change old threat-based predictions.
This is often why people can feel unsafe in a safe environment (prediction dominates) and why somatic therapy using new sensory input can help shift our nervous system state.
Trauma tightens the loop and reduces flexibility
When someone has experienced trauma their predictions of danger can become rigid or fixed and there is and orientation to cues of danger vs cues of safety. A persons body sensations can also feel intense and overhwlmeing or they might feel an absence of sensation, more numb or a sense of an internal ‘void’ or nothingness. In this state our nervous system loses the ability to update itself flexibly. She need new pieces of information to be able to interrupt what has become fixed. New sensory information can help to soften those predictions, and drop in new information to help the body and brain have a felt sense of safety, to know it’s safe now, and that nothing bad is happening in this moment.
This model validates both somatic and cognitive approaches. Somatic interventions change the input and cognitive and relational work can change the predictions. Both are required for long term sustainable change.
How Sensory Receptors Work: How the Body Turns Experience Into Signals That Regulate the Brain
To understand why somatic therapies work, we need a clear sense of how the body converts physical events, pressure, movement, temperature, breath, visceral change into electrical signals that reach the brain. This process is not mysterious, it is a well-established physiological mechanism mentioned earlier, known as mechanotransduction. Every sensory receptor in the body is constantly transforming physical change into neural activity that directly influences autonomic regulation and emotional state.
Mechanotransduction: How sensation becomes electricity
Mechanotransduction is the process by which receptors open ion channels in response to stretch, pressure or chemical change. When these channels open:
The receptor generates a graded receptor potential ( the electrical response is proportional to the amount of mechanical force, eg more pressure or stress =larger potential and vice versa.
If strong enough, this becomes an action potential (An action potential is all-or-nothing, it either happens fully or not at all). A rapid electrical spike that travels along the axon. That signal carries sensory information to the spinal cord and brain along afferent fibres
This process of your body is constantly sending electrical updates to your brain about your internal and external world is described in:
Nature Reviews Neuroscience (Delmas & Coste, 2013) and Neuron (Abraira & Ginty, 2013).
Different receptors carry different kinds of regulatory information
Mechanoreceptors (touch, pressure, stretch)
Located in skin, fascia and muscle, they detect grounding, weight, contact, and movement. They shape autonomic tone through signals ascending via dorsal column pathways into brainstem regulation centres. Safe, slow touch changes autonomic state.
Read about: Slow, rhythmic pressure reduces amygdala activity and increases parasympathetic engagement (McGlone et al., 2014).
McGlone et al. (2014) CT-touch increases parasympathetic tone and reduces sympathetic activation. Morrison (2016) Affective touch supports emotion regulation and decreases stress reactivity.
Proprioceptors (muscle spindles, Golgi tendon organs)
These receptors monitor muscle length and tension. They send constant updates to the cerebellum and reticular formation areas that regulate posture, balance and arousal. Read about: Proprioceptive feedback stabilises autonomic output and reduces sympathetic activation (Proske & Gandevia, 2012).
Proske & Gandevia (2012) Proprioception stabilises autonomic output through brainstem pathways.
Berntson et al. (1998) Somatic afferents interact directly with autonomic regulation systems.
Baroreceptors (heart–blood pressure sensors)
Found in the carotid sinus and aortic arch, baroreceptors detect each heartbeat and communicate with the nucleus tractus solitarius (NTS).
When baroreceptors fire more frequently (for example during slow exhalation), sympathetic activity decreases and parasympathetic dominance increases.
Read about: Increased baroreceptor firing through slow exhalation lowers sympathetic drive and influences threat perception (Garfinkel et al., 2014).
Garfinkel et al. (2014) Baroreceptor activation reduces threat perception and amygdala reactivity.
Benarroch (1993) Baroreflex is a primary regulator of autonomic output.
Interoceptive C-fibres (visceral sensation)
C-fibre interoceptive pathways communicate visceral states (e.g., breath, gut sensations, cardiac signals) to the insula, creating the core of feeling states. These slow, unmyelinated fibres report from the gut, lungs and organs. They project to the posterior insula, forming the foundation of feeling and emotional tone
Read about: Interoceptive input to the insula is central to emotional regulation (Craig, 2002).
Craig (2002) Interoceptive signals form the representational basis of emotion, affect and self-awareness.
Khalsa et al. (2018) Interoception is central to emotion regulation and mental health.
Vagal afferents (body-to-brain via vagus nerve)
The vagus nerve is made up of 80% afferent fibres which means that it is carrying data from our body about heart, lungs and viscera upward towards the brain.
Read about: Vagal afferent stimulation modulates inflammation, mood and autonomic reactivity (Breit et al., 2018).
Agostoni et al. (1957) Foundational fibre count studies.
Breit et al. (2018) Vagal afferents influence mood, inflammation and autonomic regulation.
These signals converge in brain regions that regulate state
Sensory input from the body flows into key regulatory hubs:
Brainstem (NTS, parabrachial nucleus) for moment-to-moment autonomic control
Insula for interoceptive awareness and emotional feeling states
Anterior cingulate cortex (ACC) for conflict detection, emotional regulation
Medial prefrontal cortex (mPFC) for top-down regulation and meaning
Amygdala/PAG for threat detection, defensive responses
This shows that sensory data directly shape emotional experience, arousal, and patterns of protection.
Sensory input is not “background noise” it is foundational to regulation
Every breath, weight shift, muscle contraction, heartbeat and touch is converted constantly into neural signals.
These signals:
Adjust heart rate
Influence threat appraisal
Determine whether the system moves toward safety or mobilisation
Give the brain the data it uses to update or correct predictions
Shape a persons moment to moment experience of self
Why this matters clinically
Understanding mechanotransduction and interoception clarifies why somatic therapies can be supportive:Gentle touch changes insular activity.
Slow breath changes baroreflex engagement.
Movement changes proprioceptive signalling.
Grounding and posture change muscle spindle output.
Vagal afferents shift inflammation, mood, and heart-rate regulation.
These are not metaphors, and they are not “soft science.” These are hardwired, measurable, physiological pathways. Somatic therapy can help us directly change the sensory information that shapes regulation.
All these forms of sensory input converge in autonomic and emotional control centres
Regardless of receptor type, all sensory information eventually funnels into:
NTS (primary autonomic centre)
Parabrachial nucleus (emotional integration)
Insula (interoceptive feeling states)
ACC (emotion regulation, conflict detection)
mPFC (meaning-making and top-down modulation)
This is where sensory signals become shifts in:
Heart rate
Breathing
Emotional tone
Threat appraisal
Behavioural readiness
The scientific consensus is clear
Across physiology, affective neuroscience, psychophysiology and interoception research, the same conclusion appears. Sensory input from the body is not optional. It is a primary driver of autonomic and emotional regulation. Somatic therapy works because it supports change of the sensory input that feeds into these regulatory systems. Cognitive and relational processes work because they change the brain’s predictions about that input. Together, they create a flexible, resilient regulatory loop.
Why Somatic Therapy Is Effective: A Synthesis of Predictive Processing, Interoception and Trauma Neuroscience
Somatic therapy is often described as “bottom-up,” but contemporary neuroscience is showing that its effectiveness comes from how it interacts with the brain’s predictive systems, interoceptive pathways, and autonomic regulation circuits. Rather than working against the brain’s predictive tendencies, somatic work uses new sensory experiences to reshape old threat-based expectations. This makes it uniquely suited to trauma, where protective patterns are held not only in our cognition, but in the body’s signalling and the brain’s habitual interpretations.
Trauma biases the brain toward threat-based predictions
Predictive processing research shows that the brain is constantly generating anticipatory models of what the body will feel. In trauma, these predictions become over-learned, rigid, and biased toward danger, even when the present moment is safe.
This results in:
Hyperarousal
Shutdown
Misinterpretation of bodily cues
Chronic tension or collapse
Difficulty sensing safety
In this state, the brain “expects” danger and interprets bodily sensations through that lens.
Somatic therapy can. change the input the brain receives
Somatic interventions introduce new patterns of interoceptive and proprioceptive data, things like:
Slow exhalation → increased baroreceptor firing → sympathetic reduction
Grounding pressure → mechanoreceptor engagement → reduced limbic activation
Gentle movement → proprioceptive modulation → stabilised arousal
CT-touch → insular activation → parasympathetic increase
Posture/orientation shifts → proprioceptive change → sense of agency
These changes are not symbolic. They are real biological shifts in the signals that reach the brain. We give the brain new data, to support it to learn something new and interrupt the prediction.
New sensory input creates prediction errors that update the brain
Prediction errors occur when the body’s sensory signals do not match the brain’s expectations. For trauma survivors, this is important, because:
The brain expects danger
Somatic interventions produce signals of safety
This mismatch forces the brain to reconsider its model
This is how somatic work drives real neuroplastic change.
Somatic therapy strengthens interoceptive accuracy and reduces “false alarms”Research on interoception demonstrates that trauma often creates blurred or distorted body awareness. Somatic therapy helps by:
Improving the clarity of internal sensation
Supporting clients to track subtle cues without overwhelm
Teaching differentiation between high arousal and actual threat
Strengthening insular pathways involved in body awareness
As we deepen our interoceptive capacity become more more able to trust and understand our internal body signals, an important part of regulation.
Somatic work restores disrupted networks involved in self and safety
As we have discussed, shutdown, freeze and dissociation involve disconnection between:
Insula
mPFC
ACC
PAG
Somatic cues help to reintegrate these networks by reintroducing safe, tolerable sensory information that the brain can process without overwhelming stress. This gently repairs the circuitry involved in threat appraisal, emotion regulation and bodily self-awareness.
Somatic therapy requires input with co-regulation
We know that human nervous systems regulate each other and most importantly we need co-regulation and the support of another human, brain, body and heart when we are working with developmental and attachment traumas like emotional neglect and abuse.
Co-regulation through our
Voice tone
Facial expression
Rhythmic timing
Proximity
Attuned presence
Are all important elements of relational safety. As much as our modalities and interventions are important parts of the work, it is also ultimately the relationship that provides safety and healing. This supportive container that allows people to experience safety and when combined with somatic therapy, co-regulation is an important part of building a persons self regulatory capacity.
Somatic therapy supports agency, movement and choice
Freeze and collapse statesman restrict voluntary movement, narrowing our person’s physiological options. Somatic interventions can help us reclaim:
Micro-movements
Gestures of protection or reaching
Shifts in posture
Orienting toward safety
Breath flexibility
The ability to move from immobilisation into action
These embodied experiences can restore a sense of empowerment, the sense of “I can” that trauma often takes away. Bringing all of this together we can see that:
Predictive processing explains how trauma shapes the brain’s expectations
Interoception explains how bodily signals construct emotional experience
Sensory neuroscience explains how mechanoreceptors, baroreceptors, proprioceptors and vagal afferents shape state
Trauma neuroscience explains how networks become disconnected and overwhelmed
Co-regulation research explains how relational cues shift physiology
Somatic therapy sits at the intersection of all these systems.
Bringing It All Together: A Modern, Integrated Model of Regulation
If neuroscience has made one thing clear it is that regulation does not come from a single system, a single circuit, or a single theory. Instead, it emerges from the dynamic interplay of prediction, sensation, autonomic responses, relational cues and meaning. This integrated perspective offers a deeper, more precise understanding of trauma and explains why somatic therapies remain essential, even as models like Polyvagal Theory are refined or critiqued.
Regulation is co created by the body and brain
Regulation is not generated solely in the cortex or the autonomic nervous system. It emerges from the ongoing negotiation between the two. The brain continually forms predictions about what the body will feel, while the body continually sends sensory data that confirm or contradict those predictions. Neither process works in isolation; they operate as a closed loop.
This means we regulate not because the brain decides to calm down and the body follows, but because the brain and body find synchrony.
Safety and danger are computed by distributed networks
Critics of Polyvagal Theory are right that there is no single “safety system” or “neuroception circuit.” Instead, threat and safety emerge from multiple interacting networks, including:
insula
amygdala
mPFC
ACC
brainstem nuclei (NTS, parabrachial complex)
periaqueductal gray (PAG)
cerebellum
These networks integrate interoception, sensory input, memory, prediction and context. This is why trauma can feel global. It involves the whole system, not just one reflex.
Bottom up sensory input plays a foundational role
Mechanoreceptors, baroreceptors, proprioceptors, interoceptive fibres and vagal afferents provide the raw data from which emotional experience is built. Decades of research have shown that:
Baroreceptor activation shifts sympathetic arousal (Garfinkel et al., 2014)
CT-touch modulates insula and reduces amygdala activity (McGlone et al., 2014)
Proprioception stabilises brainstem arousal circuits (Proske & Gandevia, 2012)
vVgal afferents influence mood, inflammation and autonomic tone (Breit et al., 2018)
This evidence base is independent of Polyvagal Theory and supports the core principle: The body has genuine regulatory power.
Top down predictions shape how the body is felt
Predictive processing shows that the brain filters, shapes and interprets bodily sensations based on prior learning. Trauma biases predictions toward danger, making neutral sensations feel threatening or overwhelming.
This explains why clients may react with fear even in safe environments. Their brain expects danger and shapes their felt experience accordingly. But predictions can change and somatic therapy provides the new sensory inputs needed to update old models.
Co-regulation amplifies safety through relational cues
Human nervous systems are social. Research on autonomic synchrony and co-regulation (Feldman, 2012; Bornstein, 2023; Helm et al., 2012) consistently shows that facial expression, tone of voice, proximity and attuned presence have measurable effects on heart rate, vagal tone and emotional state. Even critics of Polyvagal Theory agree on this point: connection regulates physiology. Somatic therapy leverages this through attuned pacing, warm presence and shared regulation.
Trauma disrupts the loop, somatic therapy can support restoration
Trauma overwhelms the system by:
Increasing sympathetic drive
Breaking the synchrony between body and brain
Distorting interoceptive signals
Stiffening the brain’s predictions
Disconnecting key networks (insula, mPFC, PAG)
Restricting movement and agency
Somatic therapy restores regulation by:
Changing the body’s sensory input
Challenging old predictions
Reconnecting disrupted neural networks
Widening the window of tolerance
Supporting co-regulation
Reintroducing movement, breath and orientation
Offering new experiences of safety
An integrated model is more accurate than any single theory
Polyvagal Theory brought essential language for understanding states of safety, mobilisation and shutdown.
Predictive processing explains how the brain interprets sensations. Interoception research clarifies where feelings come from.
Trauma neuroscience highlights the network-level disruptions underlying freeze, collapse and dissociation.
None of these models contradict one another. Together, they provide a fuller, richer, more precise map of human regulation.
The Future of Trauma Therapy Is Integrative, Embodied and Evidence Informed
As neuroscience advances, it is becoming increasingly clear that no single theory, whether Polyvagal, predictive coding, interoception research, or trauma network models can fully explain the complexity of regulation or the brain. Each offers a piece of the puzzle, and together they reveal a truth that is both scientifically rigorous and deeply aligned with clinical wisdom. Regulation emerges from the continuous, reciprocal dance between body, brain, and relationship.
Somatic therapies remain important because they change the input the brain receives. We can create changes through breath, movement, posture, interoception and vagal feedback in ways that shift autonomic state and challenge old threat-based predictions. Predictive processing models highlight how trauma shapes the brain’s expectations, making bodily signals feel dangerous even when the environment is safe. Interoception research shows how sensations become feelings and forms the foundation of emotional life. Trauma neuroscience demonstrates that shutdown, freeze and dissociation arise from distributed networks across the insula, mPFC, ACC and PAG networks that can be gently reorganised through sensory, relational and embodied experience. Relational science confirms that co-regulation which includes things like facial expression, voice tone, proximity and attuned presence has measurable, biological effects on our nervous system.
I believe the critiques of Polyvagal Theory do not diminish the value of somatic work they refine and add to iy. They remind us to move beyond simplistic descriptions and toward a more nuanced, multi network understanding of safety and danger. They ask us to honour the complexity of human physiology without losing the clinical clarity that helps clients make sense of their experience.
What emerges is a model of regulation that is not top-down or bottom-up, but more connected, or looped a living system in which prediction and sensation, body and meaning, self and other, continually shape one another. This integrative perspective invites a more precise, compassionate and flexible approach to therapy. I am hopeful it is one that respects each individuals unique expereince, that recognises the protective brilliance of survival responses, and offers new experiences of safety that support neural change.