In the last article, I traced a chain from a single gene to a curved spine. POC5 builds cilia in the inner ear. Mutate it, and the vestibular hardware that detects gravity is malformed. The sensing tilts. The spine follows. LBX1 builds the proprioceptive interneurons that feel where the trunk is in space. Impair it, and the body cannot detect the asymmetry as it develops.
That article was important because it proved something: scoliosis genes do not code for bone. They code for sensing. The spine is downstream.
But genetics is only one door into the building.
If you read that article and thought, “I don’t have a genetic mutation, so this doesn’t apply to me,” stay here. This is the article where the model opens up. Because the mechanism that produces a scoliotic curve is broader than any single cause. It is a loop. And there are many ways in.
What causes scoliosis if it’s not genetic?
The short answer: scoliosis develops when the nervous system’s map of the body falls behind the body’s growth. The sensory systems that track where the spine is, where vertical is, how the trunk is oriented in space, lose accuracy. The body schema becomes outdated. And an outdated map generates asymmetric motor output, which generates an asymmetric spine.
This is the neural generation hypothesis.
> “The neural generation hypothesis proposes that scoliosis develops when sensory degradation in the vestibular, proprioceptive, or visual systems creates a rate mismatch between nervous system mapping and skeletal growth, locking the spine into a self-reinforcing curve attractor loop.”
The genetic evidence from the previous article is proof of concept for this model. It showed one traceable chain: gene > sensing organ > degraded signal > asymmetric output > curve. But the model does not require a gene to activate. It requires degraded sensory input during growth. And there are many ways to degrade sensory input.
Genetics is one. Trauma is another. Developmental issues are another. Chronic postural habits are another. Vestibular damage, proprioceptive deficits, visual dysfunction. All of them converge on the same mechanism. Different doors, same room.
The rate mismatch: why adolescence is the danger zone
Here is the concept that ties everything together.
During adolescent growth spurts, the vertebral column grows fast. In some kids, staggeringly fast. The spine can add centimeters in months. But the nervous system that maps and manages that spine does not grow at the same rate.
The spinal cord, the vestibulospinal pathways, the proprioceptive networks that track trunk position, all of these develop on their own timeline. This is what researchers call “uncoupled neuro-osseous growth.” The bones and the nerves are not synchronized. They are running on different clocks.
In most adolescents, this mismatch is small enough that the nervous system compensates. It updates its body schema frequently enough, accurately enough, to keep pace with the skeletal changes. The map stays close enough to the territory.
But if the sensory systems feeding that map are already compromised, already degraded by genetics, by trauma, by vestibular dysfunction, by anything, then the nervous system cannot keep up. The map falls behind. The body schema becomes inaccurate.
And an inaccurate body schema does not just fail to correct asymmetry. It actively generates it.
The nervous system organizes muscle activation based on its model of where the body is. If the model says the body is straight when it is not, the motor output will maintain the curve. If the model says vertical is tilted five degrees to the left, every righting reflex, every postural correction, every unconscious adjustment will organize around that tilted reference.
The curve is not a failure of structure. It is the faithful output of a map that no longer matches the territory.
The curve attractor loop
The rate mismatch explains how the curve starts. The curve attractor loop explains why it accelerates.
Once a small asymmetry begins, a cascade locks it in place. Here is how it works.
Asymmetric loading changes bone growth. This is the Hueter-Volkmann law, one of the most established principles in orthopedic biomechanics. When one side of a vertebral growth plate is compressed more than the other, the compressed side grows slower. The tension side grows faster. Asymmetric muscle activation produces asymmetric spinal loading. Asymmetric loading, sustained through a growth spurt, produces vertebral wedging. The bone itself begins to encode the curve.
The curve degrades proprioception further. As the spine curves, the proprioceptive map of the trunk loses resolution. This is cortical smudging. The brain’s representation of the curved region becomes blurry, less differentiated. The person literally cannot feel the asymmetry as clearly as someone with a straight spine can feel their symmetry. The sensor that should detect the problem is degraded by the problem itself.
The body schema accepts the new shape as baseline. The gamma efferent system, the neural mechanism that sets muscle tension baselines, recalibrates to the curved position. The asymmetric contraction pattern that is holding the curve becomes the new default. The muscles are not “tight” or “weak” in the conventional sense. They are following an updated set point. The nervous system has decided this is normal.
The pattern becomes invisible to the person living in it. This is why people with scoliosis often say they feel straight. They are not in denial. Their body schema genuinely reports straightness because the schema itself has been updated to match the curve. The measuring instrument has been recalibrated to the wrong standard.
The loop feeds itself. Degraded sensing produces asymmetric output. Asymmetric output produces asymmetric bone growth. Asymmetric bone growth further degrades sensing. The sensing degradation prevents detection and correction. The curve deepens. The attractor strengthens.
This is not a metaphor. This is a dynamical systems description of a self-reinforcing biological process. In the language of nonlinear dynamics, the curve is an attractor state. The basin of attraction gets deeper with every cycle of the loop. The longer the loop runs, the harder it is to escape.
Why this explains what nothing else does
The curve attractor model answers questions that the standard structural model cannot.
Why are 80% of cases “idiopathic”? Because the cause is the rate mismatch, and the rate mismatch is invisible on X-ray. You cannot see a degraded body schema on a Cobb angle measurement. You cannot see a vestibular asymmetry on a spinal film. The diagnostic model tracks the shape, not the generator. The cause looks unknown because the tools are pointed at the wrong layer.
Why do curves progress during growth and stabilize after skeletal maturity? Because the rate mismatch resolves when growth stops. Once the vertebral column is no longer changing faster than the nervous system can map it, the primary driver of curve progression disappears. The curve stabilizes not because the problem was fixed, but because the window of vulnerability closed. The attractor is still there. It just stops deepening.
Why are girls affected 8:1 at severe curve magnitudes? Girls enter puberty earlier and have a longer, more gradual growth period. The window during which the rate mismatch can operate is wider. There is also emerging evidence that hormonal factors affect proprioceptive sensitivity, potentially degrading the sensing layer during the exact period when it needs to be sharpest. The gender disparity is not about bone weakness. It is about timing and sensing.
Why does early intervention work better? Because the attractor is shallower early on. The loop has not had time to deepen. The bone has not yet permanently wedged. The cortical smudging is not yet severe. The gamma efferent recalibration is not yet fully consolidated. Interrupt the loop early, and there is less to undo. Wait, and the attractor basin deepens with every growth cycle.
Why does “idiopathic” persist as a label? Because the standard diagnostic model was built to describe shapes, not to identify generators. “Idiopathic” is not a mystery of the body. It is a limitation of the framework.
The many doors into the same room
The genetic entry point was covered in detail in the previous article. But the neural generation hypothesis identifies at least six entry points into the curve attractor loop. All of them degrade the sensory input that feeds the body schema. All of them can initiate the rate mismatch.
Genetic. POC5 and LBX1 degrade vestibular and proprioceptive hardware from development. The sensing organs are structurally impaired before any curve begins. This was the focus of the previous article because it provides the most measurable, traceable chain. You can sequence the gene. You can image the malformed vestibular canal. You can trace the full path from mutation to sensing deficit to asymmetric output to curve.
Vestibular. Any asymmetry in the otolith organs tilts the gravitational reference frame. This does not require a genetic mutation. Developmental variation in semicircular canal geometry, birth trauma affecting the temporal bone, chronic inner ear infections during critical developmental periods, all of these can create a vestibular bias.
The animal evidence is definitive. Lambert and Straka (2009) removed vestibular organs from one side of frog larvae. The frogs developed scoliosis. Full lateral curvature, sagittal deformation, vertebral rotation. The pattern matched human scoliosis. The pinealectomy studies in chickens showed the same thing from a different angle: disrupt vestibular processing through melatonin removal, and 75% develop scoliosis. Only in bipedal species. Because bipedal posture depends on accurate gravity sensing in a way that quadrupedal posture does not.
Proprioceptive. Research consistently shows that adolescents with idiopathic scoliosis have measurably impaired trunk position sense compared to controls. They cannot feel where their trunk is in space as accurately. This is not a consequence of having a curved spine. Studies measuring proprioceptive accuracy independently of curve severity show that the deficit exists as a primary feature, not just a secondary effect.
VEMP studies show that 38% of AIS patients have abnormal vestibular evoked responses. Subjective visual vertical testing shows they perceive gravitational vertical inaccurately. These are not structural findings. They are sensing findings. And they precede and predict the curve.
Trauma. This is the entry point that the standard model ignores entirely. Thomas Hanna documented the Trauma Reflex: a protective lateral flexion pattern that is triggered by injury, surgery, sustained emotional threat, or any significant physical insult. The reflex pulls one hip up, shifts the trunk laterally, and rotates the ribcage. It is mediated by the tectospinal tract and maintained by the gamma efferent system.
If the reflex is never resolved, it consolidates. The nervous system treats the contraction as normal. The asymmetric muscle activation becomes the baseline. During an adolescent growth spurt, this chronic asymmetric activation feeds directly into the Hueter-Volkmann mechanism. The bones grow into the shape that the trauma reflex is holding.
The threat cascade that trauma initiates does not just affect emotions. It reorganizes the physical structure. This is one of the most common and least recognized entry points into scoliosis.
Developmental. Retained primitive reflexes, specifically the Asymmetric Tonic Neck Reflex (ATNR) and the Tonic Labyrinthine Reflex (TLR), create asymmetric postural patterns that should have integrated in infancy. When they persist, they impose a bias on the motor output that the developing nervous system organizes around. During rapid growth, that bias becomes the seed of a curve.
Nutritional. Vitamin D deficiency is correlated with scoliosis, and the mechanism may be dual. Vitamin D affects bone density, obviously, making the skeleton more susceptible to deformation under asymmetric loading. But it also affects brain morphology and neural development. The sensing and the structure are both compromised simultaneously. This is an area where the research is still emerging, but the convergence is suggestive.
All roads lead to the same place: degraded sensory input. Body schema mismatch. Asymmetric motor output. Asymmetric loading. Curve.
The entry point varies. The mechanism is the same.
Why the genetic evidence matters even if you don’t have the gene
In the previous article, I focused on POC5 and LBX1 because they provided the cleanest proof. You can sequence a genome. You can image a vestibular canal. You can measure the structural asymmetry in the sensing organ. You can trace the chain from mutation to malformed cilia to degraded gravitational signal to asymmetric vestibulospinal output to curve.
That chain is measurable at every link. That is what makes it powerful as evidence.
The other entry points are harder to measure. You cannot easily quantify the degree to which a trauma reflex has degraded trunk proprioception. You cannot put a number on how much a retained ATNR is biasing motor output during a growth spurt. The logic is the same, but the measurement tools are less precise.
The genetic evidence is proof of concept for the entire model. It demonstrates that the causal chain runs through sensing, not structure. It proves the mechanism. The other entry points use the same mechanism. They just enter through different doors.
This is why “is scoliosis genetic?” is both the right question and the wrong question. Yes, genetics is one entry point. No, it is not the only one. The deeper question is: what is the mechanism that all the entry points share? And the answer is: sensory degradation creating a rate mismatch during growth, locking the spine into a curve attractor loop.
What my 85 degrees taught me about this model
I was diagnosed at sixteen with an 85-degree kyphoscoliosis. Nobody tested my vestibular function. Nobody assessed my proprioceptive accuracy. Nobody asked about trauma history. Nobody checked for retained reflexes.
They measured the curve. They called it idiopathic. They recommended surgery.
I spent fifteen years in the structural model, trying to fix a shape with shape-based interventions. Exercises for the spine. Stretches for the muscles. Corrections aimed at the output.
None of it lasted because none of it addressed the generator. The loop kept running. The attractor kept pulling.
When I finally found the upstream layer, the sensing layer, things shifted that had not shifted in a decade and a half. Two inches of height gained. Not from surgery. Not from bracing. From updating the system that was generating the curve.
The vestibular recalibration. The proprioceptive restoration. The resolution of bracing patterns that had been running since adolescence. The body schema updated because it finally received accurate information. And the structure followed the map.
I do not tell this story to suggest that everyone’s curve will respond the same way. I tell it because it demonstrates the model. The curve had a generator. The generator had an address. When I found the address, the system could reorganize.
The curve has a generator. The generator has an address.
If you have scoliosis and you have been told the cause is unknown, you have not been given a wrong answer. You have been given an incomplete model.
The neural generation hypothesis does not reduce scoliosis to one cause. It identifies one mechanism that multiple causes feed into. Genetics, vestibular dysfunction, proprioceptive deficit, trauma, developmental retention, nutritional factors. These are different entry points into the same loop. The loop is the curve attractor. And the curve attractor is what keeps the spine where it is.
The standard model sees the curve and asks how big it is. The neural generation model sees the curve and asks what is generating it. Those are fundamentally different questions, and they lead to fundamentally different answers about what is possible.
The curve has a generator. The generator has an address. We know where to find it.
If you are ready to look upstream from the shape to the source, our scoliosis-specific assessment identifies your entry point into the attractor loop and builds a protocol around the generator, not the curve.
Start with a scoliosis-specific assessment at Posture Dojo.
—
Sam Miller is the founder of Posture Dojo and the author of UPRIGHT. He was diagnosed with 85-degree kyphoscoliosis at age 16 and spent 15 years in the standard model before finding the generator. He found it.
—
Sources
– Lambert FM, Malinvaud D, Glaunes J, Bergot C, Straka H, Vidal PP. “Vestibular asymmetry as the cause of idiopathic scoliosis: a possible answer from Xenopus.” J Neuroscience. 2009;29(40):12477-83. – Lambert FM, Straka H. “Restricted neural plasticity in vestibulospinal pathways after unilateral labyrinthectomy as the origin for scoliotic deformations.” J Neuroscience. 2013;33(16):6845. – Machida M et al. “Characterization of scoliosis after pinealectomy in the chicken.” Spine. 1997. – Machida M et al. “Production of scoliosis after pinealectomy in chickens, rats, and hamsters.” Spine. 1998. – Kucerova K et al. “Vestibular functions of adolescents with idiopathic scoliosis: a comprehensive assessment.” J Manipulative Physiol Ther. 2025. – Cakrt O et al. “Subjective visual vertical in patients with idiopathic scoliosis.” J Vestib Res. 2011;21(3):181-6. – Simoneau M et al. “Sensory reweighting is altered in adolescent patients with scoliosis: evidence from a neuromechanical model.” Gait Posture. 2015;43:164-169. – Wang D et al. “Sensory integration dysfunction in adolescent idiopathic scoliosis.” J Clin Neurosci. 2025;134:111267. – Shi L et al. “Lateral semicircular canal asymmetry in idiopathic scoliosis.” PLoS One. 2015;10(7):e0131120. – Hanna T. Somatics: Reawakening the Mind’s Control of Movement, Flexibility, and Health. Da Capo Press, 1988. – Defined G et al. “Vestibular morphological alterations in adolescent idiopathic scoliosis: a systematic review.” Symmetry. 2023;15(1):175. – Cyr JP et al. “Enhanced vestibular-evoked balance responses in adolescents with idiopathic scoliosis.” J Neurophysiol. 2025.