Is the Polyvagal Theory Dangerous? (2) It Is Time to Take a Critical Look at the Criticism
On Evidence, Double Standards and the Ethics of Criticism
Lees je liever de Nederlandse versie?
Last week, the first part of this two-part series was published, in which the experiences of clients and professionals with the polyvagal theory took centre stage. In this new article, I expose a pattern. The polyvagal theory (PVT) may be subjected to critical scrutiny; that is both valuable and necessary. But the burden of proof that some demand of it stands in stark contrast to the standards applied elsewhere in science. That inconsistency deserves attention because it gives the impression that more is at play than purely scientific motives1.
To make that pattern visible, I do not begin with the PVT but with a number of medical and psychological treatments that have been standard practice for decades. Because anyone who wants to understand the bar being set for the PVT must first see how low that bar lies elsewhere2. Many treatments applied daily in clinics and practices rest on a foundation that is, on closer inspection, rather shaky. Treatments that work are sometimes only proven to do so years later, and treatments that appeared to work are sometimes only found to be ineffective years down the line.
Let me first give a few examples of how low the bar in science sometimes lies. The contrast with what critics demand of the PVT will then be all the sharper.
Paracetamol: still a mystery
Paracetamol is the most widely used painkiller in the world. You can buy it at any pharmacy without a prescription. And yet scientists still do not know precisely how it works.
What we do know: paracetamol does not work like most other painkillers. Ibuprofen and aspirin reduce inflammation by targeting a specific process in the body. Paracetamol barely does this at all. How it relieves pain remains incompletely understood. There are indications that a breakdown product of paracetamol, formed in the brain, acts on the pain system via pathways that normally respond to cannabis-like substances in the body (Högestätt et al., 2005; Mallet et al., 2023). But even that has not been proven. Multiple processes are likely at work simultaneously.
No one demands that paracetamol be withdrawn from the market until its mechanism is fully elucidated, and that would be absurd; the drug helps people, that much is beyond doubt. But remember this example. I will return to it.
Lithium: gold standard without a golden explanation
For people with bipolar disorder, in which moods can swing violently between deep depression and mania, lithium is considered one of the most effective medications available. It is also used in recurrent depression. There is strong evidence for its role in stabilizing mood and reducing the risk of suicide (Cipriani et al., 2013). But what does lithium actually do in the brain? That is less clear. It affects dozens of processes in the body simultaneously. Which of those processes explains its efficacy remains a matter of scientific debate (Malhi et al., 2013). We prescribe it. It works, but we do not know precisely how. No problem (apparently).
EMDR: are the eye movements actually necessary?
EMDR is a well-researched therapy for trauma and is now included in clinical guidelines worldwide. That it works is not in dispute. But the specific element to which EMDR owes its name, the eye movements, very much is. When researchers compared EMDR with eye movements to EMDR without eye movements, an initial large-scale analysis found no difference in effectiveness (Davidson & Parker, 2001). A later analysis did find a modest difference (Lee & Cuijpers, 2013), but that study contained a number of weaknesses that make its conclusions less conclusive. A broad overview of 76 studies concluded that the evidence for the specific contribution of eye movements is not unequivocal (Cuijpers et al., 2020). The most widely supported explanation now points to working memory load: simultaneously holding a memory and performing a mental task reduces the emotional charge of that memory. This can be achieved by means other than eye movements.
Does EMDR work? Yes. But not for the reason embedded in its name. And yet it appears in every trauma guideline.
Attachment theory: useful, influential, and less proven than it appears
Bowlby’s attachment theory is one of the most influential models in twentieth-century psychology. The core idea is recognizable and intuitive: the bond a child forms with its caregiver(s) in the early years lays a foundation for how that person later navigates relationships, handles stress, and regulates themselves. Therapists worldwide work with it. Clients recognize themselves in it. It gives language to something that is otherwise difficult to articulate.
But the scientific status of attachment theory is shakier than its broad acceptance suggests. The original categories—secure, anxious-avoidant, anxious-ambivalent, and disorganized—are based on the Strange Situation, a laboratory procedure from 1969 in which a toddler is separated from its caregiver for a few minutes (Ainsworth & Wittig, 1969). The question is, how representative is such a brief laboratory situation of the full complexity of a caregiving relationship? And how well does what you measure in that moment actually predict how someone will manage relationships as an adult? The answer: less well than you might expect (Fraley, 2002). Whether early attachment actually causes later difficulties, rather than simply being associated with them, has not been convincingly demonstrated3.
And then there is the oversimplification. In popular psychology, attachment theory has become an explanatory system capable of accounting for almost anything: relationship anxiety, avoidance, burnout, addiction. That is not what Bowlby described, and it does not do the theory justice. A useful framework becomes a catch-all concept and, in the process, becomes simultaneously less useful and harder to test.
Familiar? It is a pattern we also see with the PVT…
Herniation and knee surgery: unnecessary?
This is where things become genuinely uncomfortable. Lumbar disc surgery was performed on a large scale for many years and still is in many countries. The reasoning is intuitively plausible: there is pressure on a nerve, we relieve that pressure, and the pain disappears. Only, on closer inspection, that story holds up only in part.
A large-scale American study, the so-called SPORT trial, compared surgery with conservative treatment, in this case physiotherapy and watchful waiting. The conclusion: in the long term, surgery barely outperformed conservative management (Weinstein et al., 2006). Moreover, it turns out that disc herniations can resolve spontaneously in many patients without any intervention. And there is something else striking: a substantial proportion of the general population has demonstrable disc herniations on MRI scans without experiencing any pain whatsoever. The structural abnormality on the scan and the patient’s complaint are only weakly correlated. The scan is used to justify the procedure, while that same scan does not adequately explain why the person is in pain.
Then there is the parallel with knee arthroscopy for osteoarthritis, a procedure in which cartilage in the knee is trimmed or cleaned. In 2002, researchers published a study in the prestigious New England Journal of Medicine in which some patients underwent a sham operation: an incision was made, but nothing was done inside the knee. After two years, these patients had the same outcomes as those who had undergone the actual procedure (Moseley et al., 2002). The operation performed no better than the expectation of having been operated on. It still took years before practice changed.
Here, then, not only is the explanation lacking; the active mechanism itself is absent. And yet the surgery continued for years.
Antidepressants and the serotonin story
For a long time, the dominant explanation for depression was the so-called serotonin deficiency hypothesis: depression was said to be caused by insufficient serotonin in the brain, and antidepressants, the so-called SSRIs, were supposed to correct that deficit. This explanation appeared in patient information leaflets, was conveyed by general practitioners, and gave people an understandable account of what was happening to them.
In 2022, a large umbrella review summarized all available research in this area. The conclusion was unambiguous: there is no consistent scientific evidence that depression is caused by a serotonin deficiency (Moncrieff et al., 2022). SSRIs help some patients; that is not in dispute. But the explanation that was proclaimed with such confidence for decades turns out to be wrong. Whether a treatment works and whether the theory behind it is correct are two different questions. And here, the theory was not correct.
A ruptured aneurysm: when the bar disappears entirely
No randomized controlled trial (the gold standard for establishing the safety and effectiveness of a treatment) has ever been conducted in patients with a ruptured abdominal aortic aneurysm (a tear in the body’s main artery, usually in the abdomen, which causes the patient to bleed to death internally unless surgery is performed). The control condition, “let’s wait and see,” has never passed an ethics committee, and it never will. And yet surgeons worldwide operate daily on patients with a rupturing aorta, without anyone arguing that this is unethical lacking RCT evidence. Clinical logic suffices: the artery ruptures, the patient bleeds to death, and we intervene. Proven mechanism? Check. Proven efficacy via randomized trial? Well, no, but no one loses sleep over it.
It is just one example. But it illustrates something: the call for evidence is not always equally loud and not always equally consistent. Sometimes clinical logic may suffice. Sometimes common sense is permitted to take precedence over protocol. It is simply a pity that this latitude is not extended with equal generosity to everyone.
The double standard and how it is maintained
Disc surgery was performed and reimbursed on a large scale for years, while evidence for its added value was already under pressure. An incorrect explanation for depression was conveyed to patients for decades. EMDR appears in every guideline, while its most distinctive element remains scientifically unsettled. No one seems particularly troubled by any of this.
A theory about the functioning of the autonomic nervous system, with a solid grounding in neuroanatomy, recognized in clinical practice, and supported by a growing body of empirical research (see, for example, Porges, 20254), must, according to some, meet stricter requirements before it can be taken seriously.
It is therefore time to subject the criticism itself to critical scrutiny. Because the way in which the PVT is judged is not arbitrary. This situation is maintained by a combination of structural mechanisms and by the content and tone of the criticism itself. I identify six aspects:
The first is publication bias. Studies with positive results are published more frequently than studies with negative results. For established treatments, this means failures remain invisible for longer. For new theories such as the polyvagal theory, it means every critical voice finds a platform, while supporting evidence is subjected to stricter scrutiny.
The second is the question of who actually decides what constitutes “sufficient proof.” Clinical guideline committees, health insurers, and journal editorial boards—all of them have their interests, blind spots, and networks. The polyvagal theory has no powerful institutional lobby behind it. Paracetamol and SSRIs do. That has consequences for which treatments and theories become mainstream and which must first prove themselves entirely5.
The third is the shifting goalposts. The knee arthroscopy example showed that a sham operation worked just as well as a real one. The effect consisted entirely of the expectation of recovery following the (sham) procedure. But when a PVT-informed therapist does something that helps, and clients recover, regulate, and feel what they had not felt for a long time, they are asked to demonstrate the mechanism, or it is said that working in this way is unethical. The standard shifts according to what is convenient.
The fourth is the language of evidence itself. “Not evidence-based” sounds like an objective judgment: neutral, scientific, beyond question. But the term is not neutral. The dominant model of scientific evidence was developed for standardized interventions: a pill, a protocol, or a procedure. That model works reasonably well for medications or surgery. It works less well for explanatory theories about physiological processes, relational safety, or the functioning of the autonomic nervous system. Theories that do not describe a single intervention but an underlying principle, such as the PVT, which does not prescribe what a practitioner should do but explains why people respond the way they do under pressure, in connection, in danger, and in safety. Those who use “evidence-based” as the sole standard without acknowledging its limitations are, from the outset, excluding part of reality. And calling that objectivity.
The fifth is the reversed burden of proof. In the debate around the PVT, the claim is regularly made that polyvagal-informed practice is harmful. But that claim has never been substantiated with evidence. No one has presented documented cases. No one has cited formal complaints. No one has provided case reports. The burden of proof for efficacy is stringently demanded, but the burden of proof for the alleged harm caused is not delivered. That is not what we call scientific reasoning; it is an unsubstantiated claim and a line of argument that does not apply to itself the very norm it insists upon. That is dishonest and perhaps even lacking in integrity.
Which brings us to the ethical dimension. In the debate, working from a polyvagal perspective is sometimes described as outright unethical because the theory “does not hold” or is insufficiently proven. That is a bold claim. A claim that, applied consistently, would also require removing paracetamol from the market, prohibiting disc surgery, and striking EMDR from clinical guidelines. No one draws that conclusion there. That raises the question of whether it is ethically defensible to assert that working with the PVT is unethical, and therefore who is actually violating an ethical norm here.
The way in which criticism is currently being expressed does not sound like science. It sounds more like politics. And those who look more closely at the tone, the consistency, and the targeted nature of the criticism of the PVT may well ask themselves whether scientific integrity is the driving force here, or something else entirely6.
Read here part 1 of this two-part series, in which I invite clients and professionals to share their experiences with the polyvagal theory, both positive and negative.
If you found this article worth reading and (not yet) feel like getting a paid subscription, you can always treat me to a cappuccino!
References
Ainsworth, M.D.S. & Wittig, B.A. (1969). Attachment and exploratory behavior of one-year-olds in a strange situation. In B.M. Foss (red.), Determinants of infant behaviour (Vol. 4, pp. 111–136). Methuen.
Cipriani, A., e.a. (2013). Lithium in the prevention of suicide in mood disorders: updated systematic review and meta-analysis. BMJ, 346, f3646. https://doi.org/10.1136/bmj.f3646
Cuijpers, P., e.a. (2020). Eye movement desensitization and reprocessing for mental health problems: a systematic review and meta-analysis. Cognitive Behaviour Therapy, 49(3), 165–180. https://doi.org/10.1080/16506073.2019.1703801
Davidson, P.R. & Parker, K.C.H. (2001). Eye movement desensitization and reprocessing (EMDR): a meta-analysis. Journal of Consulting and Clinical Psychology, 69(2), 305–316. https://doi.org/10.1037/0022-006X.69.2.305
Fraley, R.C. (2002). Attachment stability from infancy to adulthood: Meta-analysis and dynamic modeling of developmental mechanisms. Personality and Social Psychology Review, 6(2), 123–151. https://doi.org/10.1207/S15327957PSPR0602_03
Grossman, P., e.a. (2026). Why the polyvagal theory is untenable. An international expert evaluation of the polyvagal theory and commentary upon Porges, S.W. (2025). Clinical Neuropsychiatry.
https://doi.org/10.36131/cnfioritieditore20260110
Högestätt, E.D., e.a. (2005). Conversion of acetaminophen to the bioactive N-acylphenolamine AM404 via fatty acid amide hydrolase-dependent arachidonic acid conjugation in the nervous system. Journal of Biological Chemistry, 280(36), 31405–31412.
https://doi.org/10.1074/jbc.M501489200
Kok, B.E. & Fredrickson, B.L. (2010). Upward spirals of the heart: Autonomic flexibility, as indexed by vagal tone, reciprocally and prospectively predicts positive emotions and social connectedness. Biological Psychology, 85, 432–436.
https://doi.org/10.1016/j.biopsycho.2010.09.005
Lee, C.W. & Cuijpers, P. (2013). A meta-analysis of the contribution of eye movements in processing emotional memories. Journal of Behavior Therapy and Experimental Psychiatry, 44(2), 231–239. https://doi.org/10.1016/j.jbtep.2012.11.001
Malhi, G.S., e.a. (2013). Potential mechanisms of action of lithium in bipolar disorder. CNS Drugs, 27(2), 135–153.
https://doi.org/10.1007/s40263-013-0039-0
Mallet, C., Desmeules, J., Pegahi, R. & Eschalier, A. (2023). An updated review on the metabolite (AM404)-mediated central mechanism of action of paracetamol (acetaminophen): experimental evidence and potential clinical impact. Journal of Pain Research, 16, 1081–1094. https://doi.org/10.2147/JPR.S393809
Moncrieff, J., e.a. (2022). The serotonin theory of depression: a systematic umbrella review of the evidence. Molecular Psychiatry, 28, 3243–3256. https://doi.org/10.1038/s41380-022-01661-0
Moseley, J.B., e.a. (2002). A controlled trial of arthroscopic surgery for osteoarthritis of the knee. New England Journal of Medicine, 347(2), 81–88. https://doi.org/10.1056/NEJMoa013259
Porges, S.W. (2026). When a critique becomes untenable: A scholarly response to Grossman et al.’s evaluation of polyvagal theory. Clinical Neuropsychiatry.
https://doi.org/10.36131/cnfioritieditore20260111
Quintana, D.S., Guastella, A.J., Outhred, T., Hickie, I.B. & Kemp, A.H. (2012). Heart rate variability is associated with emotion recognition: Direct evidence for a relationship between the autonomic nervous system and social cognition. International Journal of Psychophysiology, 86(2), 168–172.
https://doi.org/10.1016/j.ijpsycho.2012.08.012
Weinstein, J.N., e.a. (2006). Surgical vs nonoperative treatment for lumbar disk herniation: the Spine Patient Outcomes Research Trial (SPORT). JAMA, 296(20), 2441–2450.
https://doi.org/10.1001/jama.296.20.2441
You might think of personal motives (perhaps some critics have a personal grievance or professional disagreement with Stephen Porges that has taken on a life of its own), but also consider the views generated when a negative article is posted with a provocative headline (“Polyvagal Theory is Debunked!”), of which I have seen several examples.
This article does not offer a blanket defense of the PVT. The neuroanatomical objections raised by Grossman et al. (2026) deserve a substantive response, and one exists: Porges has responded to them at length (Porges, 2026). That debate is not revisited here. What this article does take a position on is the standard to which the PVT is held. Anyone who wishes to criticize the PVT is entitled to do so, but on the same grounds applied to paracetamol, disc surgery, and EMDR. No stricter or more selective than those. Simply consistent.
This is a common misunderstanding in psychology and the popular literature: the confusion of correlation with causation. Two things that are associated do not necessarily cause one another. Children with insecure attachment more frequently experience difficulties in relationships or stress regulation in later life; that finding is consistent. But whether early attachment causes those difficulties, or whether both are caused by a third factor, such as temperament, genetic predisposition, or broader family circumstances, is a separate question. That question has rarely been definitively answered. Correlation says these two things go together. Causation says one causes the other. That is a fundamental distinction, and it is not always carefully observed in the attachment literature or, indeed, in a great many areas of science.
A few examples of empirical support: research strongly suggested that people with higher baseline heart rate variability experienced higher levels of positive emotions and displayed a pattern of reciprocal effects between positive emotions, social engagement, and vagal tone (Kok & Fredrickson, 2010). Another study directly tested the PVT prediction that autonomic regulation is associated with the capacity to recognize social signals and found empirical support for that hypothesis (Quintana et al., 2012).
Alongside the example of the ruptured aortic aneurysm, a comparable pattern can be observed with influenza vaccination: double-blind randomized research is considered ethically problematic there because the control group would be deliberately exposed to an avoidable risk and because the vaccine is assumed to be ‘proven effective.’ But that assumption is less solid than it appears. Effectiveness varies considerably from year to year and averages somewhere between 40% and 60%. Cochrane reviews have questioned the robustness of the evidence, particularly for healthy adults. Without an RCT, we cannot truly be certain, yet the RCT is withheld for precisely that reason. Who decides when evidence is ‘good enough’ to render further research ethically impermissible? And who decides when that same argument does not apply?
See Grossman et al. (2026) and the response by Porges (2026).