Incidence of Cranial Asymmetry in Healthy Newborns

MATERIALS AND METHODS

The study was composed of 201 essentially healthy neonates at Brigham and Women’s Hospital who were examined 24 to 72 hours after delivery. We excluded infants with a gestational age <36 weeks and infants who required prolonged treatment in the intensive care unit. After obtaining consent for participation, parents were questioned about the pregnancy and birth history. Data were collected on gender, height, birth weight, gestational age, Apgar score, and possible medical problems and anomalies. Also documented were number of pregnancies and deliveries; possible uterine or pelvic anomalies; and events during the pregnancy (oligo- or polyhydramnios, early pelvic descent, and other complications). Information was also noted about mode of delivery, use of forceps or vacuum assistance, length of first- and second-stage labor, position at birth, and birth complications. Medical records of the neonate and the mother were reviewed for all of these items.

Physical examination of the neonate noted cranial or facial asymmetry, head shape, cephalohematoma, torticollis, and anomalies of the ears, the hands, the feet, and the hips. Height was measured in the crib, using a measuring board, for most infants; however, in some infants, such as those recently circumcised, height and hip stability were recorded from the charts.

Anthropometric cranial measurements were made, using a spreading caliper (Tyrostom Company, Olomouc, Czech Republic), with the head held in a neutral position by an assistant or the mother. Two oblique cranial diameters were determined, measuring from the supraorbital point (os, orbitale superius) to the parietooccipital scalp at the point of maximal convexity, as described previously.⁵ These diameters were measured 3 times, the average was recorded in centimeters and used to calculate the transcranial difference (TCD). Measurements of the head circumference were performed 3 times using a measuring tape; the average was noted in centimeters. All measurements were made by the same person with the same measuring device.

If an infant had either localized head flattening or cranial asymmetry, defined as a TCD of >4 mm, the parents were encouraged to change the infant’s position in the crib frequently and to position the infant away from the flattened area. Parents were also offered the opportunity to have their infant’s head shape reevaluated after 2 or 3 months in our craniofacial unit. The parents were given the transverse cranial diameters to allow comparison of the initial and the 2-month follow-up measurements without identifying the child in the primary study.

Data were tested for statistical significance with Pearson χ² tests or with 1-way analysis of variance.

RESULTS

Of the total of 201 neonates, 183 were singletons and 18 were twins from 9 pairs. Singletons and twins were analyzed separately.

Head Shapes

Of the 24 infants with flat areas, 9 were right posterior, 4 right lateral, 9 left posterior, 1 left lateral, and 1 bilateral posterior flattening. Thus, right-sided head flattening accounted for 54.2% of the total of the flattened areas. In most infants, the flat area was well localized, between 5- and 7-cm diameter. Six neonates with head flattening had anterior displacement of the ipsilateral ear (Fig 1).

• Download figure
• Open in new tab
• Download powerpoint

Fig 1.

Photograph of head of patient 90, a female born by cesarean section for breech position to a primipara. Head evidenced right posterior-lateral flattening, with measurements R os L scalp: 11.4 cm and L os-R scalp: 10.6 cm, for TCD: 8 mm. Infant had R ear prominent and minor R torticollis.

In the group of 21 infants with otherwise unusual head shapes, 11 were found to have cephalohematoma; 7 presented with dolicocephalus or a protruding occiput; and 2 had molding in the region of the vertex. Two neonates had bilateral frontal flattening, 1 of these also had a cephalohematoma.

DISCUSSION

This study suggests that congenital, localized flattening of the occiput or the posterior parietal cranium is a precursor of posterior deformational plagiocephaly. Many of the findings in neonates with localized flat areas are the same as in infants with fully expressed posterior deformational plagiocephaly diagnosed at a later age:

1. Boys are more commonly affected. In our study, we found a male versus female ratio of 2:1. In infants with posterior deformational plagiocephaly, predominance of males is uniformly reported, with gender ratios between 3:1⁵ and 3:2.³¹ Possible explanations include that males have larger heads than females and that male fetuses are less flexible than females, which makes them more susceptible to deformational anomalies.²⁹ We did not find that head circumference differed between infants with normal and flat heads. This discrepancy is probably explained by our small numbers.

2. Cranial flattening is more frequently observed on the right side. In our study, right-sided flattening was present in 54%, left sided flattening only in 41%, corresponding to observations in infants with deformational plagiocephaly.^5,11 One possible reason for the higher incidence of right-sided flattening might be the preponderance of the left occipital anterior presentation at birth. In this position, the infant’s right occiput is compressed against the maternal pelvic bone and the left forehead against the lumbosacral spine. Also postnatally, the infant’s head is rarely straight in the midline, and infants turn their head to the right side in nearly 80%, according to Volpe.³² Thus, preponderance of right-sided flattening is not surprising.³³

3. We assumed that, similar to infants with fully expressed deformational plagiocephaly, neonates with localized head flattening have an increased risk of having other deformational anomalies. Our study showed a significant association between cranial flattening and auricular deformations, both commonly observed in neonates. We found other deformational anomalies as well, such as hip dysplasia, club feet, and mandibular hypoplasia, but our population was too small to make a definite statement about a statistically significant correlation. With regard to torticollis, there is an unresolved controversy over which is the primary disorder: cranial asymmetry or torticollis, or are both congenital anomalies.^1,2,33–35 In our craniofacial unit, we observed torticollis in 64% of infants with frontal deformational plagiocephaly⁷ and in 26% of children with posterior deformational posterior plagiocephaly.⁵ In this study, however, torticollis was only noted in 2 infants (<10%), both of whom had ipsilateral flat areas. Thus, the prevalence of torticollis was much lower in neonates with precursor deformation than in infants with fully developed deformational plagiocephaly. Our study supports the belief that, frequently, torticollis associated with head deformation develops postnatally. Thus, in most cases, torticollis is not the primary anomaly causing cranial asymmetry; rather stiffness of the neck muscles develops secondary to deformational plagiocephaly if the head is permanently tilted toward the “position of comfort.” It could equally be argued that neonatal torticollis is a causative factor and often missed in routine examination. We relied not only on the pediatricians, but also our study examiner was alert to the potential finding.

4. Cephalohematoma is a well-known risk factor for posterior deformational plagiocephaly.²⁹ In infants with localized cranial flattening at birth, we found a much higher incidence of cephalohematoma, compared with all other newborns. If infants with a localized flat area also have a contralateral cephalohematoma, they are even more likely to lie preferentially on the flat area. Thus, they are at very high risk of progression to full deformational plagiocephaly.

5. Both multiple birth and intrauterine constraint have been reported as risk factors for fully expressed deformational plagiocephaly.^30,36 We observed a prevalence of 56% of head flattening in twin births, which is presumably a reflection of intrauterine constraint. The effects of this constraint are discussed controversially. Some authors suggest that compressive and restrictive intrauterine forces cause secondary conformational changes³⁰ or even fusion of cranial sutures.^36–39 We think that intrauterine constraint acts as a molding force and does not cause true sutural synostosis.⁵ This is consistent with the observation that crib positioning or a molding helmet can easily correct cranial flattening.

6. Localized flat areas are more likely to occur with primiparity, assisted delivery, and long labor. All of these conditions have also been identified as risk factors for deformational plagiocephaly³⁶ and for other deformational anomalies. It is obvious that they cause increased molding, acting on the fetal head during delivery.

We emphasize that a well-localized flat area we observed in our study does not constitute fully developed “neonatal plagiocephaly.” However, we suggest that many of these infants with posterior or lateral flattening will likely develop posterior deformational plagiocephaly. In the past, when infants were positioned prone, this common posterior flattening improved spontaneously. The infant with congenital posterior flattening lying supine will prefer to turn the head to the flat side because this position is most comfortable. Thus, congenital occipitoparietal deformation is perpetuated and accentuated by pressure from the mattress. Assuming this pathogenesis, the term “positional” rather than “deformational” plagiocephaly has some merit, although the precursor to deformation is present at birth. As this flattening is quite subtle, it is frequently overlooked in the neonatal period. This explains why most mothers of infants with deformational plagiocephaly recall that their infant had a normal head shape after birth,³⁶ and why parents and pediatricians often do not notice deformational plagiocephaly until 2 to 3 months of age.⁵

This study showed a prevalence of 13% of localized cranial flattening in essentially healthy newborn singletons. The incidence of fully developed posterior deformational plagiocephaly in infancy is not known, as there have been no good prospective studies.⁴⁰ Data from the 1970s suggest an incidence of 1:300 for occipital plagiocephaly.⁴¹ Given the almost epidemic increase in the problem reported by many craniofacial centers since the 1992 recommendation of the American Academy of Pediatrics, the present incidence is likely much higher. Furthermore, without a prospective study, we cannot predict how many neonates with localized flat areas will develop full deformational plagiocephaly. Nor can we know whether the 13% of newborns with congenital flat areas account for all of the increased incidence. The reasons that we did not do this study are: 1) this would require funding to follow 200 healthy infants; 2) and, more importantly, we had alerted the parents to the cranial flattening and had given instruction to keep the infant’s head off the flat area. Therefore, such a study would be scientifically flawed.

Neonates with a gestational age of <36 weeks and neonates who required prolonged intensive care treatment during their first days of life where excluded from our study. Preterm infants, low birth weight infants and seriously compromised neonates, are, however, more likely to have anomalous head shapes than essentially healthy newborns.⁴² All pediatricians and neonatologists are familiar with the dolicocephalic head so commonly observed in preterm infants; often called “preemie head.” Preterm infants have malleable calvarial bones,^43–46 making them more susceptible to molding forces during birth. In addition, newborns who receive prolonged treatment in the intensive care unit have a higher incidence of neurologic impairment, which is another risk factor for deformational cranial anomalies.⁴⁷ Thus, we assume that the true incidence of cranial flattening in all newborn singletons is slightly higher than in the population assessed in this study.

All examined newborns were classified as having either cranial flattening, infants with an unusual head shape, or normal head shape, according to our physical examination. Because the measurements were recorded within 24 to 72 hours of life, it could be argued that the localized flattening we found is a transient phenomenon. However, in our experience, cranial molding improves within hours of birth and resolves within days. Hence, our finding of localized flattening on the second and third days of life is unlikely to be evanescent.

When the anthropometric measurements for these 3 groups were compared, there was, all in all, a very good correlation between transcranial measurements and physical findings. The mean TCD in infants with a normal head shape was <1 mm and a TCD of 4 mm was, in general, perceived as a posterior flat area. However, a few infants with TCDs of 4 mm had a perfectly normal head shape on physical examination. This is a general problem of anthropometry: the measured difference does not always correlate with the perceived degree of asymmetry.

The few infants whose heads looked completely symmetrical but with a measured TCD of 4 mm should be carefully monitored. However, decisions about treatment should be made according to the perceivable degree of asymmetry, and not based solely on anthropometry.

CONCLUSION

Our study showed that localized areas of cranial flattening can be identified in a significant percentage of grossly healthy neonates. We consider these flat areas to be precursors to posterior deformational (“positional”) plagiocephaly. Progression to full deformational plagiocephaly is likely to occur if such an infant continuously lies on the flat side of the head. Thus, we recommend the following:

1. Neonates with occipital flat areas can be easily identified. Parents should be on alert for cranial flattening. If it occurs, the parents should change position in the crib regularly and be certain the infant does not always lie on the flat side of the head.

2. Amendation be made to the recommendations of the American Academy of Pediatrics, acknowledging that the incidence of SIDS has dropped significantly with the supine sleep position. When lying on the back, the infant’s head should be turned regularly. As an alternative, the infant could be positioned on either side, with a blanket behind the back. The parents should be instructed to allow supervised “tummy time” while their infant is awake. They should also be told to be cautious about the amount of time their infant spends in a car seat.

These are simple instructions that we consider effective to avoid progression from a localized flattened cranial area to posterior-lateral deformational plagiocephaly. We believe that if both parents and pediatricians paid more attention to the infant’s head shape and crib positioning, the high incidence in posterior deformational plagiocephaly would decrease significantly.