Interoceptive inference in POTS: the miscalibrated system
A 2018 paper by Owens and colleagues in Frontiers in Physiology applies predictive processing theory — one of the most influential frameworks in contemporary neuroscience — directly to the physiology of POTS and dysautonomia. The argument is precise and consequential: dysautonomia is not a broken autonomic system. It is a miscalibrated one. That distinction changes what treatment should target.
What Predictive Processing Means for Autonomic Regulation
Predictive processing is the neuroscientific framework in which the brain does not passively receive sensory information. It actively generates predictions about what signals it expects from the body, then updates those predictions based on incoming data. When predictions match reality, the system is stable. When they don't, the brain issues corrective signals to close the gap.
Interoceptive inference is this same process applied to internal body signals. The brain maintains a running model of heart rate, blood pressure, blood volume, breathing rate, and visceral state. It predicts what those signals should be at any given moment. It compares those predictions against what it actually receives. The difference — the prediction error — drives the autonomic output: the adjustments to vascular tone, heart rate, respiration, and sweat response that keep physiology within tolerable range.
In a healthy, well-calibrated system, this process runs smoothly. The brain's model is accurate enough that its predictions closely match reality, prediction errors are small, and the corrective signals are appropriately sized. In dysautonomia, the model has drifted. The brain is generating predictions about internal body state that do not match what the body is actually doing. The corrective signals it issues in response are themselves inaccurate — too large, too small, or aimed at the wrong target. The tachycardia, the vasomotor instability, the hypersensitivity to postural change: these are not evidence of a broken peripheral nervous system. They are the outputs of a central control system operating from a faulty model of the body it governs.
Why "Miscalibrated" Changes the Clinical Question
The difference between broken and miscalibrated is not semantic. A broken system needs to be repaired or replaced. A miscalibrated system needs to be recalibrated — and the levers for doing that are different from the levers for managing symptoms.
Standard dysautonomia treatment largely focuses on suppressing the outputs of the miscalibrated system: beta-blockers reduce the tachycardia, volume loading compensates for the perceived hypovolemia, compression garments push against peripheral pooling. These interventions can be genuinely useful. But Owens and colleagues are pointing to something they cannot do: they do not update the brain's internal model. The model continues generating inaccurate predictions. The system continues issuing corrective responses calibrated to those predictions. The treatment is managing the output of a miscalibrated process rather than correcting the miscalibration itself.
The interoceptive inference framework raises a specific question that symptom-management approaches do not: what is causing the brain's predictive model of internal body state to be wrong? Possibilities include: insufficient or noisy afferent signal quality from the periphery (so the brain cannot update its model accurately), learned maladaptive priors built from prior episodes of physiological disruption, or central processing abnormalities that amplify prediction errors beyond what the incoming signal warrants. Each of these has different implications for what would actually recalibrate the system.
Interoceptive Accuracy, Signal Noise, and the Feedback Loop
One mechanism the paper highlights is interoceptive accuracy — the degree to which a person can correctly perceive their own internal signals. Research in interoception has consistently found that heightened symptom reporting in functional conditions correlates not with peripheral pathology but with abnormal central processing of interoceptive signals: either amplified prediction errors, imprecise internal models, or both.
In POTS, this creates a potential feedback loop. Orthostatic stress generates real physiological changes — cerebral blood flow drops before heart rate rises, and the body's corrective responses escalate. The brain's model of what those signals mean is being updated in real time. If the model is already miscalibrated — overweighting certain signals, underweighting others — then each orthostatic episode reinforces the miscalibration rather than correcting it. The system is not stabilizing around a new setpoint. It is iteratively updating toward a worse model.
This does not make the symptoms imagined or functional in the pejorative sense. It locates the mechanism more precisely: in the central processing of real physiological signals, rather than exclusively in the peripheral generation of those signals.
What the Research Cannot Tell Us (Yet) and What It Can
The Owens paper is a theoretical framework piece, not a clinical trial. It does not demonstrate that interventions targeting interoceptive recalibration produce better outcomes than symptom-management approaches. What it does is provide a mechanistic account of dysautonomia that is consistent with several clinical observations that symptom-management models struggle to explain: why symptom severity and measurable physiological abnormality often diverge; why similar objective findings produce wildly different symptom experiences across patients; why some patients improve substantially with interventions that do not directly address their measured autonomic abnormalities.
The framework is also not an argument against volume loading, compression, or rate control. It is an argument for asking a second question alongside the standard one. Standard question: how do we manage the outputs of this dysregulated system? Second question: what is maintaining the miscalibration, and can it be directly addressed?
What This Means for the Reader
If you have POTS or dysautonomia and the treatments you have tried have only partially worked, the predictive processing framework offers one explanation: they were addressing the outputs of the miscalibration rather than the miscalibration itself. Your autonomic system is not broken in the sense of being irreparably damaged. It is operating from an inaccurate model of your body's internal state, issuing corrective responses calibrated to that inaccurate model. The question worth asking is not only how to suppress those responses, but what would allow the model to become more accurate. That is a different question than the ones most evaluations are designed to answer — and it is one this paper suggests is worth asking explicitly.