Torticollis in Children

Congenital muscular torticollis (CMT) is unilateral shortening or fibrosis of the sternocleidomastoid (SCM). The head tilts toward the affected side and the chin rotates to the opposite side. It is the most common cause of torticollis in infants and young children.

Who. CMT affects 0.4-2% of live births and is the third most common congenital musculoskeletal condition, after DDH and clubfoot. The right side is affected more often (75%), and males slightly more often than females. It is associated with difficult delivery, breech presentation and first-born infants.

Why. The most accepted theory is an intrauterine compartment syndrome of the SCM. Others propose birth trauma with ischaemia of the muscle, or venous occlusion and oedema leading to fibrosis. Oligohydramnios and multiple pregnancy are also associated.

Not every tilt is muscular. CMT accounts for 80% of paediatric torticollis. The remaining 20% includes osseous abnormalities (Klippel-Feil, atlantoaxial rotatory subluxation), tumours (posterior fossa, spinal cord), ocular causes and Sandifer syndrome, and the features that should send you looking for them are listed under Investigations.

The muscle. The sternal head of the SCM arises from the manubrium and the clavicular head from the medial clavicle; both insert on the mastoid process and the lateral superior nuchal line. Surgical release is usually at the clavicular end and rarely proximally. The muscle is supplied by the spinal accessory nerve (CN XI), which is at risk in surgery.

The deformity. The SCM tilts the head to its own side and rotates it to the opposite side, which explains the clinical pattern of CMT. A unilaterally shortened muscle holds the head in its own action: tilted toward the tight side, chin turned away.

View larger

Secondary deformities. Left untreated, the tilt remodels the skull and face, and these changes are what drive the urgency of early treatment:

• Plagiocephaly - flattening of the ipsilateral occiput, the side the infant lies on
• Facial asymmetry - the ipsilateral face appears smaller
• Frontal bossing - in positional plagiocephaly the forehead on the side of the flat occiput comes forward (see the pearl below)

Timing. Positional plagiocephaly develops within months if torticollis is untreated. Mild cases improve with treatment, but severe plagiocephaly may require helmet therapy, which is why early stretching matters: it prevents the secondary skull deformity.

CMT is described by what is felt in the SCM and graded by how much rotation is lost. All three clinical subtypes generally respond well to conservative treatment if it is started early.

The pattern. The head tilts toward the tight SCM and the chin points away from it: a tight right SCM gives a right tilt and a chin pointing left. Passive rotation of the chin toward the affected side is limited, and in early infancy there may be a palpable pseudotumour (fibromatosis colli) in the muscle.

History. Ask about the birth (breech, prolonged labour, forceps or vacuum), when the tilt was first noticed and whether it is worsening, stable or improving, and any stretching or physiotherapy already tried. Feeding difficulty may point to Sandifer syndrome and visual concerns to an ocular torticollis.

Examination. Work from the head down to the hips and feet:

• Inspection - tilt direction and chin position, plagiocephaly (flattening of the occiput), facial asymmetry (ipsilateral face smaller), and neck webbing, which may indicate Klippel-Feil
• SCM palpation - a mass (pseudotumour) or tightness
• Range of motion - passive rotation should be symmetrical, so measure the deficit; passive lateral flexion to each side; active range observed as the child looks around. Asymmetry between sides is diagnostic
• Associated conditions - hips (Barlow, Ortolani, asymmetry); feet (metatarsus adductus, clubfoot); spine (scoliosis, congenital anomalies); dysmorphic features if the picture is atypical

Most CMT needs no imaging. Routine CMT with classic features is diagnosed clinically; imaging is for the child who does not fit the pattern.

When to X-ray the cervical spine. Atypical features, no palpable SCM mass, failure to respond to stretching, restricted range of motion in multiple planes, or concern for a bony anomaly.

When to consider a non-CMT cause.

• No palpable SCM mass or tightness
• Restriction in more than one plane
• Onset after the first weeks of life
• Neurological signs
• Failure to improve with proper stretching

The single most important exam skill is separating benign CMT from sinister or non-muscular causes. The atypical features listed under Investigations mandate further work-up; this table sorts the causes they point to.

Benign paroxysmal torticollis of infancy (BPTI). A torticollis that is episodic and recurrent rather than fixed points away from CMT toward BPTI. Recurrent, self-limiting attacks of head tilt (lasting hours to days) begin in the first year, often with pallor, irritability, vomiting or ataxia, and the child is entirely normal between episodes, with full passive range of motion and no SCM mass.

What it is. BPTI is regarded as a migraine variant or channelopathy (associated with CACNA1A, the gene of familial hemiplegic migraine and episodic ataxia type 2), usually with a family history of migraine. It is a clinical diagnosis of exclusion that resolves spontaneously by age 2-3 years; the priority is to exclude posterior fossa pathology and atlantoaxial rotatory fixation before reassuring.

Stretching first. Passive stretching is first-line treatment, and early, consistent stretching succeeds in over 90% when started before the first birthday. The intensity steps up with age and with failure to progress.

Algorithm

View larger

Counselling. Families should be counselled that some residual facial asymmetry may remain even after successful treatment.

After release. Post-operative physiotherapy is essential for maintaining correction, and post-operative stretching and bracing are crucial, with close physiotherapy follow-up for 3-6 months.

After conservative treatment. Stretching continues until full, symmetrical range of motion is achieved, then its frequency is weaned gradually. Positioning strategies continue through infancy, and the child is watched for recurrence.

Conservative. The best outcomes follow early diagnosis and treatment, and range of motion normalises within months in most cases.

Surgical. Series report high rates of good to excellent results, with full or near-full range of motion typically achieved; good outcomes are reported in greater than 90% when surgery is needed and done appropriately. Facial asymmetry and plagiocephaly improve but may not fully resolve, and earlier correction gives greater remodelling potential.

Global epidemiology:

• Reported birth incidence ranges from about 0.4% to as high as 16% depending on definition and screening intensity; a commonly cited clinical figure is 1-2% of newborns
• Third most common congenital musculoskeletal condition after DDH and clubfoot
• Right-sided predominance (roughly 75%); modest male predominance in several series
• Risk factors converge on intrauterine crowding/positioning: breech presentation, primiparity, oligohydramnios, multiple gestation and difficult/instrumented delivery
• The CMT-DDH association is real but smaller than the textbook 20%: verified series report around 8% (Walsh) and about 4.5% requiring treatment (Minihane), within a literature range of 2-29%

Side-by-side guidance (where emphasis differs):

How strong is the DDH association? Textbooks quote 20%, but the best primary series (Walsh approximately 8%; Minihane 4.5% requiring treatment) suggest the figure is considerably lower and that a normal clinical hip examination reliably excludes significant DDH. Whether universal hip ultrasound is justified in every CMT infant, or only in those with an abnormal examination or risk factors, remains debated.

When to operate. Conventional teaching favours release by 12-18 months, yet Shim's older children did at least as well for movement and scar (see Management). There is no high-level trial defining the ideal age.

Helmets. Evidence on whether a cranial orthosis outperforms repositioning for positional plagiocephaly is mixed. Lam reported greater measured head-shape correction with orthoses, but a randomised trial (the Dutch HEADS study, not cited here) found no clinically meaningful difference for mild-to-moderate cases. Practice varies widely by region and payer.

Open versus endoscopic release. Endoscopic or arthroscopic release (Wang) and open unipolar and bipolar techniques all report good outcomes, but no adequately powered head-to-head trial establishes superiority. The choice remains surgeon- and resource-dependent.