Neuromuscular Scoliosis

An evidence-based guide to why neuromuscular scoliosis develops, how it behaves, and how it is treated, written to give families a clear, honest, and authoritative understanding of the condition and the decisions ahead.

Neuromuscular scoliosis is not idiopathic scoliosis with a different label. It is a fundamentally different disease, and the distinction matters.

In idiopathic scoliosis, the spine curves for reasons we do not fully understand, in an otherwise healthy child, and the curve generally stops progressing once growth ends. Neuromuscular scoliosis is the direct mechanical consequence of a brain, spinal cord, nerve, or muscle that can no longer hold the trunk balanced and upright. Asymmetric muscle pull, whether from spasticity or from weakness, loads one side of each growing vertebra harder than the other. By the Hueter-Volkmann principle, the compressed side of the growth plate grows more slowly than the unloaded side, and the curve feeds on itself as the child grows.

Two facts follow from this, and they shape every decision:

The curve tracks the severity of the underlying condition. The more global the neurological disability, the higher the risk and the larger the curve. And these curves do not respect skeletal maturity. Unlike idiopathic curves, they frequently keep progressing into adult life. This is not a deformity that waits, and it is not one that bracing can reliably stop.

Neuromuscular scoliosis is an umbrella over a group of distinct diseases, each with its own behaviour, risks, and surgical considerations. Understanding the specific diagnosis is the first step in understanding the spine.

Cerebral palsy, a permanent and non-progressive injury to the developing brain, is the single most common cause of neuromuscular scoliosis seen in clinic. Across studies, scoliosis is reported in roughly a fifth to two-thirds of children with cerebral palsy, and the spread in that range is explained almost entirely by one variable: the severity of motor disability.

That severity is graded with the Gross Motor Function Classification System (GMFCS), a five-level scale where level I describes a child who walks with only subtle limitations and level V describes a child with profoundly limited self-mobility even with assistive technology. The relationship between GMFCS level and spinal deformity is one of the most robust findings in the field. In a landmark total-population study of 666 children with cerebral palsy, children at GMFCS levels IV and V carried roughly a 50% risk of moderate or severe scoliosis by the age of 18, while children at GMFCS levels I and II had almost no excess risk over the general population. Statistically, the risk rose steeply with each level, and children at the most severe level had a many-fold higher hazard of progressive scoliosis than ambulant children.

A crucial and often-misunderstood point: the type of cerebral palsy, whether spastic, dyskinetic, or another, is not an independent risk factor. The apparent link between certain subtypes and scoliosis disappears once you account for GMFCS level. In other words, it is the degree of functional disability, not the label of the subtype, that predicts the spine. This is why credible surveillance programmes are built around a child's GMFCS level and age rather than their diagnosis alone, and why curves are most often first diagnosed after about 8 years of age.

The curves themselves come in two recognisable patterns. Ambulant children tend to develop balanced double curves with little pelvic tilt. Non-ambulant, more severely affected children tend to develop a single, sweeping C-shaped curve, larger in magnitude and almost always accompanied by pelvic obliquity, a tilting of the pelvis that drives much of the disability that follows. In adults with established curves, progression of roughly 3 to 4 degrees per year is typical, and larger curves (beyond 50 degrees) progress nearly twice as fast as smaller ones.

Duchenne muscular dystrophy is a progressive, X-linked muscle-wasting disease that almost exclusively affects boys. Its relationship to scoliosis is tied closely to walking: as long as a boy remains ambulant, the spine is usually protected, but once he becomes wheelchair-dependent, scoliosis tends to follow, historically in the great majority of boys.

The modern picture is more nuanced, because routine glucocorticoid (steroid) treatment has changed the natural history, prolonging walking and delaying or reducing the severity of spinal curves in many boys. Even so, surveillance of the spine through the transition to full-time wheelchair use remains essential. Two considerations dominate surgical planning in Duchenne dystrophy more than in any other group: the heart and the lungs. Progressive cardiomyopathy and declining respiratory muscle strength mean there is often a finite window in which surgery can be performed safely, and timing decisions are made jointly with cardiology and respiratory teams rather than by the spine surgeon alone.

Spinal muscular atrophy (SMA) is an autosomal recessive disease caused by loss of the SMN1 gene, leading to degeneration of the motor neurons in the spinal cord and progressive muscle weakness. It is conventionally divided into types 0 through 4 by age of onset and the highest motor milestone achieved, from the most severe infantile forms, in which a child never sits, through milder forms compatible with walking and survival into adult life. A closely related "backup" gene, SMN2, partly modifies severity: more copies generally mean a milder phenotype.

For the spine, the rule is stark: scoliosis develops in almost every non-ambulant child with SMA, and it tends to appear early and progress relentlessly. Spinal fusion is the definitive treatment of choice. Bracing is contentious. There is no convincing evidence it halts progression, and a brace that compresses an already weakened chest wall can actively harm breathing, so it is used cautiously if at all. When severe curves appear in very young children, growth-friendly techniques (including growing rods and expandable titanium rib constructs) can hold the deformity and protect the developing chest until the child is old enough, ideally past about age 10, for definitive fusion.

SMA is also the condition in which the wider treatment landscape has changed most dramatically. A decade ago, the best the literature could offer was the prospect of antisense oligonucleotide therapy in clinical trials. Disease-modifying treatments aimed at restoring SMN protein have since moved from trial to clinic, altering the natural history for many children. These therapies do not remove the need for spinal surgery in those who have already developed significant curves, but they have reshaped expectations, and any modern discussion of SMA should take place alongside a neuromuscular specialist familiar with the current options.

In myelomeningocele, the most significant form of spina bifida, the spinal cord and its coverings fail to form properly before birth. More than 80% of children with this condition develop a significant spinal curve, and the surgical calculus here carries a unique and serious caveat that families must understand.

Extending a spinal fusion down to the pelvis, frequently necessary to correct pelvic obliquity, can cost a child the ability to walk. In high-level myelomeningocele, fusion to the pelvis has been associated with permanent loss of ambulation in anywhere from a quarter to three-quarters of patients, particularly in those who walk for exercise rather than for primary mobility. The reason is mechanical: these children often rely on a mobile lumbosacral junction to swing their legs forward, and fusing it away removes that strategy. This single fact can change the entire surgical plan, and it must be weighed honestly against the natural tendency of many of these children to lose walking ability during adolescence regardless of surgery. Two further features sharpen the picture. The skin over the back and buttocks is often insensate, raising the risk of pressure sores from both braces and prominent implants, and infection rates in this group are among the highest in all of spinal surgery.

When a child sustains a spinal cord injury, the resulting muscle weakness or paralysis removes the balanced support the trunk depends on, and a paralytic scoliosis frequently follows. The younger the child at the time of injury, and therefore the more growth remaining, the higher the risk and the more aggressive the curve tends to be. These children warrant close spinal surveillance for years after the injury.

A number of less common diagnoses round out the group. Rett syndrome, a neurodevelopmental disorder seen mainly in girls, is associated with an early and often rapidly progressive scoliosis. Charcot-Marie-Tooth disease, an inherited disorder of the peripheral nerves, and Friedreich's ataxia, which affects coordination and balance, can each produce neuromuscular curves with their own tempo and pattern. The governing principle holds across all of them: the spine follows the nerve and muscle, and the curve must be read in the context of the whole condition.

In neuromuscular scoliosis, the deformity is rarely the whole problem. The downstream consequences are what most affect a child's life, and they are the reason surgery is so often justified.

Pelvic obliquity is central. A tilted pelvis concentrates pressure on one side when a child sits, producing the pressure points and skin breakdown that can be both painful and dangerous, and it undermines the stable, hands-free sitting that lets a child use their arms to eat, play, learn, and communicate. Large curves crowd the chest and compromise breathing, raising the risk of pneumonia and respiratory infection. Reflux and gastrointestinal difficulty are common in this population and compound the respiratory risk through aspiration. Add pain, fatigue, and the steady erosion of independence, and the case for intervention is rarely about the angle on an X-ray alone. It is about protecting function, comfort, and care.

Sound management begins with sound surveillance, and the evidence points to a clear strategy: monitor based on the child's GMFCS level (or equivalent functional severity) and age, not on diagnosis alone. A child at high functional risk is examined regularly so that a curve is caught while it is still flexible and manageable, because the outcome of surgery correlates strongly with the size of the curve at the time of treatment.

Assessment is deliberately broad. Full-length spinal radiographs are taken in the sitting or standing position, with flexibility (traction) films used to judge how correctable a curve is. MRI is used where indicated to look inside the spinal cord for abnormalities such as a syrinx or a tethered cord. Pulmonary function testing quantifies respiratory reserve in children able to cooperate. And because these are multi-system diseases, evaluation runs through a multidisciplinary team, typically a deformity surgeon, paediatrician, respiratory and cardiac specialists, anaesthetist, and physiotherapist, long before any operation is contemplated.

In a meaningful number of these children, the curve is not the whole story. There is a problem inside the spinal canal itself, involving the cord or the base of the brain, and it can be the hidden engine driving the deformity. Finding it is not optional, because it changes how, and sometimes whether, the spine is operated on.

Chiari malformation. The lowest part of the brain (the cerebellum) sits too low and presses down through the opening at the base of the skull into the top of the spinal canal. This can disturb how the cord works and is strongly associated with scoliosis, particularly when it produces a syrinx. In selected children, relieving that pressure first, through a decompression performed by a neurosurgeon, allows the curve to stabilise.

Tethered cord. Normally the spinal cord hangs freely. A tethered cord is abnormally anchored to the surrounding tissue, so it is stretched as the child grows. That traction can trigger or accelerate a curve, and it often announces itself through new leg weakness or changes in bladder and bowel function. Releasing the tether may halt the decline and, in younger children, slow the curve.

Split cord malformation (diastematomyelia). The cord is divided into two halves, sometimes separated by a spur of bone or fibrous tissue. It is typically present from birth and frequently accompanies a congenital curve. It must be identified before any correction, because straightening a spine that contains a bony septum, without first addressing it, risks injuring the cord.

Syrinx. A fluid-filled cavity within the cord itself, often seen alongside a Chiari malformation, and a recognised driver of scoliosis in its own right.

The common thread is straightforward: correcting a curve stretches and shifts the spinal cord. If the cord is tethered, split, compressed by a Chiari malformation, or thinned by a syrinx, that same correction can turn from beneficial to dangerous. This is why MRI matters, and why the threshold for it is low, particularly with a rapidly progressing curve, an atypical or left-sided curve, new neurological change, or skin markers over the spine such as a dimple, a tuft of hair, or a fatty lump. These anomalies are especially common in spina bifida, and when one is active a neurosurgeon may address it before or alongside the deformity surgery. That decision is made jointly, child by child.

Non-surgical treatment has a genuine role, but it has limits. It supports; it does not cure.

For small, flexible curves, careful observation with regular review is appropriate. Physiotherapy maintains joint range and manages tightness. Thoughtful seating, including custom-moulded supports, tilt-in-space chairs, and lateral support pads, meaningfully improves posture and comfort, and a well-designed three-point seating system has been shown to improve trunk symmetry substantially in non-ambulant children. Bracing can aid sitting balance and free the upper limbs, but the evidence is consistent and important: in neuromuscular curves, bracing does not reliably stop progression. It is, at best, a way to buy time or improve sitting, not a substitute for definitive treatment, and it must be used with great care in children who cannot feel pressure on their skin.

Two adjuncts deserve mention. Botulinum toxin can reduce limb spasticity and has been used around surgery without worsening curves. Intrathecal baclofen pumps are effective for severe global spasticity and can ease pain and care, but they sit within a genuine controversy: some series have reported curve progression after pump insertion while matched studies have not, and a pump in place at the time of spinal fusion has been linked in some reports to higher infection and re-operation rates. None of this rules out their use. It argues for careful, individualised decisions.

The goal of surgery is simple to state and demanding to achieve: a balanced spine sitting squarely over a level pelvis, with the head centred and stable sitting restored. Everything in the operation serves that aim.

The modern workhorse is posterior spinal fusion using pedicle screw instrumentation, which provides strong, three-dimensional anchors placed into the vertebrae and connected by rods, holding the corrected spine while the bone fuses into a single solid column. This is the product of a long evolution, from the early Harrington rod, through Luque sublaminar wiring and the Galveston method of pelvic anchorage, to the unit rod, and finally to today's segmental pedicle screw constructs. Contemporary series report curve corrections in the region of two-thirds and pelvic-obliquity corrections of around 70%. One large series using unit-rod instrumentation, for example, reported a 68% reduction in curve and a 71% reduction in pelvic obliquity.

Anterior release remains an important option for the most demanding curves. In this approach the spine is reached from the front, through the chest or the abdomen, the discs at the apex of the curve are removed, and the spine is loosened so that a stiff, rigid deformity can be corrected and a solid fusion secured. It is reserved chiefly for three situations: large, rigid curves that will not correct adequately from the back alone, severe pelvic obliquity that cannot otherwise be levelled, and, in skeletally immature children, the prevention of the so-called crankshaft phenomenon, in which a spine fused only from behind keeps twisting because the front of the vertebrae continues to grow. The trade-off is genuine, because reaching the spine through a body cavity adds to the scale of the operation and its risks. For that reason, and because modern posterior-only pedicle screw constructs are now so powerful, the need for anterior release has fallen considerably and it is used selectively rather than routinely. It is worth noting that in experienced hands the complication rate of a combined anterior and posterior approach has not always proved higher than posterior surgery alone.

For children who develop severe curves while still very young and growing, complete fusion is the wrong answer, because it would arrest the growth of the spine and chest and harm lung development. For them, growth-friendly constructs support and partially correct the spine while allowing continued growth, deferring definitive fusion until the child is older.

Where there is significant, fixed pelvic obliquity, most often in non-ambulant children, the rods must usually be anchored into the pelvis to level it and create a stable base. This is not a trivial extension of the operation: carrying the fusion to the pelvis increases operative time, blood loss, and infection risk, so it is done when it is needed, not by default. As a general guide, a fixed pelvic obliquity under about 15 degrees can often be left uninstrumented.

Many techniques exist, and a surgeon should be fluent in more than one. Modern practice has increasingly favoured low-profile options such as sacral-alar-iliac (SAI) screws, which sit considerably deeper beneath the skin than older iliac fixation. That is an important advantage in thin or insensate children at risk of skin breakdown over prominent hardware, and these screws align naturally with the spinal screws above. The earlier Galveston, iliac-screw, and S-rod techniques each retain a role in particular anatomies, the last being especially relevant in myelomeningocele. And here the ambulation caveat returns with full force: in a walking child, and above all in high-level spina bifida, the decision to fuse to the pelvis is weighed with great care, because it can change whether that child walks again.

The difference between a good outcome and a dangerous one in this population is very often decided before the first incision. These are multi-system diseases, and the preoperative work-up is correspondingly thorough.

On the neurological front, anti-seizure medications must be reviewed. Several alter calcium handling and reduce bone density, which affects how securely screws can be anchored, and one in particular, sodium valproate, interferes with clotting and has been associated with a roughly 26% increase in blood loss during these operations. Clotting is therefore checked carefully before, during, and after surgery.

On the respiratory front, formal lung function testing guides planning, and preoperative training in non-invasive ventilation can strengthen the respiratory muscles and smooth recovery. Modern anaesthetic and perioperative care has moved away from routine prolonged ventilation after surgery, but respiratory reserve must be known in advance.

On the gastrointestinal and nutritional front, reflux is present in up to 70% of children with cerebral palsy, raising the risk of aspiration on top of limited respiratory reserve, and poor nutrition genuinely worsens outcomes. A low preoperative albumin correlates with more complications and longer hospital stays. Nutritional status is assessed and, where necessary, improved before surgery proceeds. Pulling these threads together is the role of the multidisciplinary team, and in this population that team is not a formality. It is the safety system.

Surgery for neuromuscular scoliosis is major surgery, and the complication rate is genuinely higher than for other forms of scoliosis. Across the literature, overall complication rates range widely, reported anywhere from roughly 17% to as high as the high-60s percent, with the highest rates in non-ambulant children and those with the largest curves. A large morbidity-and-mortality database analysis put the complication rate for neuromuscular scoliosis at about 17.9%, against 10.6% for congenital and 6.3% for idiopathic scoliosis, a useful benchmark for just how much more demanding these cases are.

The specific risks worth understanding:

Respiratory complications are the most common category, such as pneumonia or a need for breathing support for extra days, reflecting these children's limited reserve.

Implant problems (rods or screws loosening or breaking, sometimes in soft bone) and infection are the next most frequent. Infection risk is highest of all in spina bifida. Best-practice prevention, including meticulous technique, intravenous antibiotics, antiseptic skin preparation, and intra-wound antibiotic powder in high-risk children, has measurably lowered these rates.

Blood loss is substantial; most children require a transfusion.

Neurological injury to the spinal cord is uncommon, and intraoperative neuromonitoring is used continuously to protect the nerves.

Pseudarthrosis, a failure of the bone to fuse solidly, can occur and may require revision. Notably, operating on very young children (under about 13) carries higher rates of both neurological complications and pseudarthrosis, one reason surgeons prefer, where the curve allows, to wait.

The most consistent finding in the literature is that parents and caregivers report high satisfaction after surgery, in some studies approaching 99% satisfied, with the large majority saying they would choose the operation again. Caregivers reliably report that their child sits better, looks more comfortable, and is easier to feed, wash, position, and care for. Sitting balance, the outcome that matters most for daily life, improves in the great majority.

But the picture has nuance. Some prospective studies have found that, while parents remain satisfied, surgery does not necessarily improve a child's underlying function or co-morbidities; one such study even found a dip in measured function in the first six months with no net gain at a year. Quality-of-life evidence, taken as a whole, supports a benefit, but the literature still lacks the large, well-controlled prospective studies that would settle every question. In short, surgery for neuromuscular scoliosis reliably improves posture, sitting, comfort, and ease of care, and is highly valued by families; it does not cure the underlying disease, and it is not primarily a way to improve function the child has already lost. It is most often about stopping a relentless deformity and protecting the abilities a child still has, and on those terms, it succeeds.

The field is moving. Best-practice infection-prevention guidelines have already lowered one of the most feared complications. For spinal muscular atrophy in particular, disease-modifying therapies have transformed what was, only a decade ago, an untreatable degeneration, and as these treatments reshape the natural history of the underlying conditions, the role and timing of spinal surgery will continue to evolve with them.

Important: This guide is intended for general education and is not a substitute for individual medical advice. Every child is different, and every decision belongs with your child and your treating team. If your child develops sudden breathing difficulty, severe pain, a rapidly worsening curve, or any abrupt change in their condition, seek medical attention promptly.

• —NICE — Low Back Pain and Sciatica in Over 16s (NG59)
• —NHS — Spinal Conditions
• —British Association of Spine Surgeons (BASS)

This information is for general patient education and is not a substitute for an individual consultation. Always seek personalised advice from a qualified spinal specialist.