Article Text

Download PDFPDF

Torticollis, facial asymmetry and plagiocephaly in normal newborns
  1. L Stellwagen1,
  2. E Hubbard1,
  3. C Chambers2,
  4. K Lyons Jones3
  1. 1
    Division of Neonatology, University of California, San Diego, Medical Center, California, USA
  2. 2
    Departments of Pediatrics and Family and Preventative Medicine, University of California, San Diego, School of Medicine, California, USA
  3. 3
    Division of Dysmorphology Teratology, University of California, San Diego, School of Medicine, California, USA
  1. Dr L Stellwagen, Mail Code 8774, Division of Neonatology, University of California, San Diego, Medical Center, 200 West Arbor Drive, San Diego, CA 92103, USA; lstellwagen{at}


Objective: To evaluate the incidence and characteristics of torticollis, plagiocephaly and facial asymmetry in normal newborn infants.

Design: 102 healthy newborn infants were examined prospectively during their birth hospitalisation for torticollis with neck range of motion (ROM) assessment and for facial, mandibular and cranial asymmetry by photographic analysis.

Results: 73% of newborns had one or more asymmetry: torticollis (16%), asymmetry of the mandible (13%), facial asymmetry (42%) and asymmetry of the head (61%). Torticollis was associated with maternal report of the fetus being “stuck” in one intrauterine position for more than 6 weeks before delivery. Moderate facial asymmetry was associated with a longer second stage of labour, forceps delivery, a bigger baby and birth trauma. Moderate cranial and mandibular asymmetries were associated with birth trauma. More than one significant asymmetry was found in 10% of newborns.

Conclusions: Asymmetries of the head and neck are very common in normal newborns, and sixteen (16%) of 102 study newborns were found to have torticollis. Such newborns, especially if they sleep supine, are thought to be at risk of developing deformational posterior plagiocephaly. Identification of affected infants may allow early implementation of positioning recommendations or physical therapy to prevent the secondary craniofacial deformations that are part of an increasingly common phenomenon.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

In 1992 the American Academy of Pediatrics and the National Institutes of Health issued the recommendation for supine sleep position in infancy to prevent sudden infant death syndrome.1 Soon thereafter, an increase in the incidence of posterior cranial deformation was noted, and the possibility that this trend was related to supine sleep position has been suggested.2 However, it has long been recognised that torticollis present at birth can lead to secondary cranial deformation.3 In fact, it is now appreciated that most infants with deformational posterior plagiocephaly (DPP) have abnormalities of their neck range of motion (ROM); however, this association is not well defined or widely accepted.47 The extent to which torticollis and other craniofacial asymmetries are present at birth is not well known, although older studies quote rates of torticollis of <1%.8 9 Nor is it well known the extent to which specific craniofacial asymmetries persist beyond the newborn period and may be amenable to therapeutic intervention. The purpose of this study was to establish estimates of the prevalence, range of severity, co-occurrence and predictors of specific craniofacial malformations and torticollis in the immediate newborn period. Establishment of these estimates can support future research evaluating the effectiveness of therapeutic interventions in the first months of life.


The study design was a prospective, cross-sectional hospital-based study of newborns born between January and June of 2004.

All parents of newly born infants available for examination were asked to participate. Parents provided written consent for the study and were questioned about maternal history before the newborn was examined. Maternal and infant medical records were reviewed for demographics, delivery details, infant size, examination and outcome. Mothers were questioned about fetal activity including whether or not the baby felt “stuck” in a position during the third trimester, and if so, how long and in which position.

Photographs of the infant were taken by LS or EH. Infants were swaddled, and their heads extended slightly for the facial photograph. An assistant then gently opened their mouths to allow visualisation of alveolar ridges for the mandibular photograph. The vertex photograph was obtained with the swaddled infants placed supine.

Neck ROM assessments were performed by placing the swaddled infants supine with the shoulders at the edge of a warming table. All neck assessments were performed by LS. Infants were calmed and positioned by an assistant. The infants’ necks were examined for the presence of a sternocleidomastoid (SCM) tumour. The infants’ heads were laterally rotated, chin to each shoulder. Neck rotation was quantitated: chin moves past shoulder (100%), to shoulder (90%), or to mid-clavicle (70%). To assess lateral neck flexion, a board with graph paper was positioned under the head. The infants’ heads were tilted ear to shoulder with gentle, steady pressure until resistance was appreciated. Sagittal suture positions were marked in neutral and at maximal lateral flexion on the graph paper. This was repeated three times on each side. Neck flexion measurements were calculated by assessing the angles from the neutral position to the maximum flexion achieved. Torticollis was defined as >15° of difference in neck mobility comparing right with left.10

Photographs were analysed by one blinded investigator (KLJ) for facial asymmetry, flattening of jaw line, elevation of eye or cheek, deviation of nose or chin point, and flattening or bulging of cranial contours. Asymmetries were visually rated qualitatively as mild, moderate or severe. Alveolar ridge photographs were assessed for mandibular inclination (jaw tilt), defined as >5° of angulation between maxilla and mandible.

Descriptive statistics were generated for means, percentages and proportions. Univariate analyses were conducted using χ2 or Fisher exact tests. All analyses were performed using SPSS V11.0 for the MAC. Approval for the study was obtained from the hospital human research protection committee.


The cohort consisted of 101 mothers and their 102 healthy newborns who had neck ROM assessment and photographic analysis during their initial birth hospitalisation. Tables 1 and 2 show the maternal and newborn study population characteristics, respectively.

Table 1 Demographic, pregnancy and delivery characteristics in 101 mothers*
Table 2 Characteristics of 102 infants*

One in three mothers related that their newborns felt “stuck” or remained in the same orientation for the last weeks of pregnancy, and were able to describe the stuck position in all but one case (table 1). Ten newborns had birth trauma (table 1): cephalohaematoma (6), clavicular fracture (1), brachial plexus injury (1), cheek abrasion (1), scalp swelling (1). One or more asymmetries was found in 73% of study newborns. A limitation of lateral flexion of the neck, or torticollis, was found in 16% of infants, with a range of 15° to 33° in difference from right to left. Lateral rotation of the neck, chin to shoulder, was normal in the subgroup of newborns in whom this measurement was performed. No subject was found to have an SCM tumour. Jaw tilt and facial asymmetry are common findings (table 3). Jaw tilt varied from 5° to 13°.

Table 3 Prevalence of torticollis, jaw tilt and facial and vertex asymmetry in 102 newborn infants

The incidence of torticollis was higher when the mother reported the baby to be stuck in a position for more than 6 weeks relative to a shorter length of time or not at all. Infants with torticollis had slightly longer body lengths than infants with symmetric neck ROM. Jaw tilt was associated with birth trauma (table 4). Moderate facial asymmetry correlated with a second stage of labour lasting longer than 60 min versus less, a forceps delivery relative to a normal vaginal delivery, the presence of birth trauma, and a larger baby. Moderate vertex asymmetry was associated with birth trauma and forceps delivery (table 5).

Table 4 Univariate analysis of measures of torticollis and abnormal jaw tilt in relation to selected maternal and infant variables
Table 5 Univariate analysis of measures of asymmetry in relation to selected maternal and infant variables

Ten percent of study newborns had more than one significant asymmetry (torticollis, jaw tilt, moderate cranial or facial asymmetry): four newborns had three different significant asymmetries and six newborns had two significant asymmetries. Figure 1 shows representative photographs of normal and affected newborns.

Figure 1 Representative infant photographs: normal (A,B), mild facial asymmetry (C), mild vertex asymmetry (D), moderate facial asymmetry (E) and moderate vertex asymmetry (F).


Although previously thought to be a relatively rare occurrence, our study indicates that most newborns have asymmetries, and as many as one in six have restricted movement of the neck. There is a spectrum in severity of torticollis and other craniofacial asymmetries, and variability in the co-occurrence of these findings. Previous studies have reported torticollis rates of 0.3–3.92%.8 11 12 However, in that most of the infants in those studies had a more severe form of neck involvement or a SCM muscle tumour, the true incidence of the broader spectrum of torticollis is likely to be much higher. Furthermore, the complex movement of the SCM muscle and the difficulty in assessing neck ROM in newborns has led to an underestimation of torticollis in infancy. Most studies have looked at lateral rotation of the neck in infants, chin to shoulder, as is done in adults.10 13 Normal newborns can laterally rotate their head well past the shoulder, 100–110° from the midline, and laterally flex their head 50–60° towards the ear. Newborns with torticollis can have some limitation of lateral rotation, but generally not preventing the chin from reaching the shoulder (90°). Lateral flexion of the neck (ear to shoulder) away from the tight neck muscle is much more likely to show restricted movement in the newborn with torticollis. We suspect that many newborns with limited neck ROM are missed because of an incomplete examination. For example, a recent study investigating risk factors for DPP failed to find limited neck ROM in 7-week-old infants with deformational plagiocephaly (newborns were not assessed because of concern about vulnerability of cervical structures). They defined normal neck ROM as lateral rotation of 90°, or lateral flexion of 30° (rather than looking at differences left to right) and found that all infants met this criterion. Interestingly, they found that positional head preference, while the baby was awake or asleep, strongly correlated with DPP at 7 weeks of age.14

Our findings support the theory that torticollis at birth is related to a constrained intrauterine position rather than delivery trauma. Torticollis was not associated with type of delivery, or any particular head presentation, but maternal recall that the infant was “stuck” in position for more than 6 weeks correlated with the presence of torticollis. Newborns with torticollis were longer than unaffected babies, but we did not find that first-borns, male gender, breech presentation or instrumented delivery were predictive as has been found by others.12 15 16 Unlike torticollis, other measures of craniofacial asymmetry correlated with a difficult birth. However, it is not clear from our data whether a traumatic, prolonged or instrumented delivery caused the asymmetry, or if the constrained fetus was more likely to have a difficult delivery.

What is already known on this topic

  • Infants may be born with asymmetries due to their intrauterine position.

  • Asymmetries from early infancy can lead to permanent deformation of the head, neck and face.

  • There is a connection between torticollis and posterior cranial deformation.

What this study adds

  • As many as one in six newborns have torticollis.

  • Most newborns have mild or moderate craniofacial asymmetries.

  • Early detection of torticollis is possible, and may present an opportunity to prevent deformational posterior plagiocephaly.

Fifty percent of study newborns had mild, and 11% had moderate, asymmetries of the vertex. Such asymmetry may be related to transient scalp swelling, haematomas, moulding, or flattening from the delivery process. However, half of the newborns with moderate cranial asymmetry also had torticollis, facial asymmetry or jaw tilt, suggesting intrauterine constraint. The cranial asymmetry that we noted at birth is relatively subtle compared with the more impressive flattening of DPP that can develop in time (fig 1).

The constrained fetus with the head tipped and the jaw pushed down on the shoulder can develop jaw tilt. We found that jaw tilt occurred in 13% of study babies, accompanied by torticollis in four cases, and facial and cranial asymmetry in two cases. Of interest with regard to prognosis, jaw tilt is thought to resolve spontaneously several months after birth, but can interfere with breast feeding, especially in those with torticollis who resist turning the head away from the tight side.17

Six percent of study newborns had moderate facial asymmetry, which correlated with birth trauma, a prolonged second stage of labour, and larger birth weight (table 5). These newborns all had other asymmetries of the neck, head or mandible. Such an infant may represent the more severely constrained fetus, occupying a fixed position that prevented proper descent into the pelvis, or with a head that is not well aligned for vaginal delivery, thus leading to a more prolonged, traumatic or instrumented birth.

Ten percent of study newborns had more than one significant asymmetry. Infants who have multiple positional deformities have been referred to by several names: the stuck baby, the moulded baby syndrome, infantile postural asymmetry, and the turned head-adducted hip–truncal curvature syndrome.9 18 19 The incidence of this syndrome is estimated at less than 1% of newborns. However, milder forms of the stuck baby syndrome are more common.

Newborns who have restricted neck ROM are at risk of developing cranial deformations that we suspect could be prevented with early identification and preventive treatment.3 13 20 If given adequate prone time when awake, most of these children should resolve the minor asymmetries they have at birth, as they did in the era of prone sleep.14 The prone, head up posture that the infant assumes during “tummy time” puts the SCM on a maximal stretch, and may be adequate treatment for mild torticollis. However, infants with very restricted neck ROM, low postural tone, or with inadequate “tummy time” may be at risk of developing DPP. Such infants may need to be referred for physical therapy or additional assessment.

The limitations of this study relate to the difficulties in accurate assessment of physical findings in newborns. Larger numbers of patients should allow us to further delineate risk factors for intrauterine constraint and to distinguish prenatal influences from intrapartum events. Standardisation of neck ROM assessment and recommendations for initiation of physical therapy would be useful for paediatric practitioners. Further research is needed to define which newborns with positional asymmetries are at risk of non-resolution of torticollis and development of DPP.


That newborns can have positional deformities as a result of their restricted intrauterine environment has long been recognised. However, these deformities are more common than previously appreciated, and may lead to further cosmetic or postural problems if not addressed properly. Providers who care for newborns should carefully assess infants for these asymmetries and provide anticipatory guidance for the prevention of cranial deformation. Further research should test the effectiveness of early interventions in newborns with torticollis or other asymmetries.


We thank Cranial Technologies Inc for providing the camera, photographic training, and study tool, Neil Finer, MD, and Marc Montminy, MD, PhD, for their thoughtful review, and Dina Campbell, RN PNP, Ali Wolf, MSN PNP, and all of the baby-holders who helped with infant assessment.



  • Funding: The study received assistance from Cranial Technologies, Inc, in the form of purchase of a camera, 2 h of photographic training, and assistance in the design and printing of the study tool.

  • Competing interests: None.

  • Ethics approval: Ethics approval was obtained from the hospital human research protection committee.

  • Patient consent: Obtained.