Eur J Pediatr (2006) 165: 158–164

DOI 10.1007/s00431-005-0027-6

ORIGINA L PA PER

Heike Philippi . Andreas Faldum . Tatjana Jung .

Holger Bergmann . Katharina Bauer .
Daniela Gross . Juergen Spranger

Patterns of postural asymmetry in infants: a standardized

video-based analysis

Received: 6 June 2005 / Accepted: 20 September 2005 / Published online: 10 November 2005
# Springer-Verlag 2005

Abstract Cervical rotation deficit (CRD) and trunk con- TC and CRD was seen in 27 infants, with a left-sided
vexity (TC) constitute the diagnosis of infantile postural convexity and left-sided head rotation deficit in two-thirds
asymmetry (IPA), which is often associated with further of the infants. Plagiocephaly was present in 27 infants,
asymmetric features. However, very little data on the entire oblique body position in 13 infants, hip dysplasia in 4
symptom complex are currently available. The aim of this infants and calcaneus foot in 11 infants. In conclusion,
study was to analyse the entire clinical spectrum of IPA infantile asymmetry pattern analysis showed that morphol-
based on a standardized video documentation. Forty-five ogical and functional anomalies are intricately linked and
infants (27 male) with an asymmetry score of ≥12 points that infants with only a single apparent sign of asymmetry
(scale: 4–24) at a median post-term age of 10 weeks (range: have actually a much more generalized disturbance.
6–16) were selected from two previously studies using
predefined criteria. CRD and TC as reactive movements to Keywords Asymmetry . Plagiocephaly . Torticollis .
an orienting head turn in the prone and supine position were Scoliosis . Posture . Infancy
assessed from video recordings by three independent
observers. Plagiocephaly, oblique body position and Abbreviations CRD: Cervical rotation deficit .
asymmetric foot position were descriptively assessed by IPA: Infantile postural asymmetry . TC: Trunk convexity .
consent of the same observers. Hip dysplasia data were TSC: Total score difference
derived from sonography charts. The assessment of the
reactive movements showed a “scoliosis” pattern in six
infants, a “torticollis” pattern in nine infants, a “mixed Introduction
prone” pattern in 13 infants and a “mixed” pattern in 26
infants. Side agreement in the prone and supine position of In infancy, muscular imbalance, unilateral neuronal dys-
function, and asymmetric skeletal disturbances may lead to
This study was supported by MAIFOR Fund of the Johannes- asymmetric features [2, 5, 6, 15, 25, 32, 33, 35]. If causes
Gutenberg University Mainz, Germany such as neuromuscular disease, cerebral infarction, plexus
paresis, hemivertebrae, or intraspinal tumours are ruled out,
H. Philippi (*)
Children’s University Hospital, the asymmetry will be defined as idiopathic and postural
Department of Paediatric Neurology, which is the most common type of infantile asymmetry [3,
Im Neuenheimer Feld 150, 6, 25]. Depending on the prominent feature of the asymmetry,
69120 Heidelberg, Germany a diagnosis of either plagiocephaly, torticollis, scoliosis, hip
e-mail: [email protected] dysplasia or foot malposition is made, and a monosymp-
Tel.: +49-6221-562337
Fax: +49-6221-565222 tomatic therapy is instituted which, in many cases, neglects
additional asymmetric features [4, 8–10, 23, 36]. In spite of
A. Faldum this and probably in the face of the “Back to sleep”
Institute for Medical Biostatistics, campaign for SID prevention and its presumably negative
Epidemiology and Informatics,
Johannes Gutenberg-University, influence on postural development, an increasing number
Mainz, Germany of studies address the entire symptom complex using names
such as “moulded baby syndrome”, “seven sign syn-
T. Jung . H. Bergmann . K. Bauer . drome”, “turned head-adducted hip-truncal curvature syn-
D. Gross . J. Spranger
Children’s University Hospital, drome”, “positional preference” or “infantile postural
Johannes Gutenberg-University, asymmetry” [5, 10, 15, 21, 22, 27, 29]. However, none of
Mainz, Germany these studies has quantitated the intricate morphological
 159

and functional relationship of the anomalies constituting the

asymmetric feature spectrum. We report here a standardized
video-based analysis of asymmetry patterns in infants with
infantile postural asymmetry which is defined as the coinci-
dence of cervical rotation deficit and trunk convexity [27].

Methods 2m
Probands From March 1999 through June 2001, 51 infants
were referred to the out-patient clinic of the University
Children’s Hospital Mainz, Germany from three different
paediatric practices for participation in a study for the
development of an asymmetry scale [27]. The paediatri-
cians were asked to refer infants with three different Fig. 1 Setting of video recording
movement patterns – symmetric, slightly asymmetric and
asymmetric – in whom the clinical course (progressive or Recording commenced with the infant’s head being held for
resolving) was uncertain. From March through to Decem- a brief period of time in the middle supine position. A head
ber 2002, 61 infants were referred from 19 paediatric turn was then induced by presenting noises, toys or the
practices in Mainz to participate in a therapeutic random- physiotherapist’s face, and moving them from one side to
ized trial evaluating the effect of osteopathic treatment on the other. After at least two turns to each side, the infant was
infantile postural asymmetry. The paediatricians were put in the prone position with its head held for a brief time in
asked to refer infants with a distinct asymmetric posture the middle prone position, and the same procedures were
which had to be treated in their opinion. For the present repeated. Cervical rotation deficit and trunk convexity were
video-based pattern analysis, 54 infants (27 male) at a assessed from video recordings by three independent,
median post-term age of 10 weeks (range: 6–16 weeks) trained observers [27]. For each item, “trunk convexity
were selected from both collectives according the following supine”, “trunk convexity prone”, “cervical rotation deficit
eligibility criteria: (1) video-based asymmetry score of supine” and “cervical rotation deficit prone”, points (range:
≥12 (mean of three independent observers); (2) gestational 1–6) with precise descriptions were assigned (Figs. 2, 3).
age ≥36 weeks; (3) absence of a neurological disease, The composition of the asymmetry scale ensures that higher
notably hemiplegia, at outset and at 10–12 months of life; scores reflect a higher degree of fixation of the asymmet-
(4) a prompt orienting response to optic and acoustic ric feature (Fig. 3). Statistical analysis indicated good
stimuli. For the evaluation of the cervical lateral flexion, reliability and consistency of the asymmetry scale with
another eight infants (5 male) with an asymmetry score an intraclass correlation coefficient of 91.5% (Cronbach
<12 points were selected from the first study collective. alpha: 0.84).
All parents provided written informed consent, and the Plagiocephaly, oblique body position and asymmetric foot
study protocol was approved by the Ethical Board of the position were assessed as descriptive categories from the
Johannes Gutenberg-University of Mainz. videos by consent of the same observers. The descriptive
categories for plagiocephaly are “left/right sided”, “occipital/
Procedures A full medical history was obtained, the infants frontal”, “focal” (only one cranial bone is affected) or
were neurologically and physically examined by a paedi- “generalized” (most often parallelogram type). Oblique body
atric neurologist and a physiotherapist and a standardized position was defined as a preferential rotation pattern of the
video of the infants in the prone and supine position was trunk to one side. Asymmetric foot position was present if
recorded [27]. All infants were properly screened for a one foot persisted in the calcaneus or adductus foot position
traumatic lesion of the sternocleidomastoid muscle, but throughout the orienting head turn to the right and left side
none was identified. For asymmetry pattern analysis, data while the position of the other foot changed. Finally, a
which fit in the asymmetry framework were extracted from
both collectives. In detail, information on intrauterine
position and from hip sonography was obtained from the
Supine
medical records. The neurological examination included Trunk convexity 1 - 6 points
the evaluation of spontaneous movements, orientation Cervical rotation deficit 1 - 6 points
responses, positional reactions and reflexology according
to Vojta [2, 12, 35].The results of the hand and foot grasp Prone
reflexes as indicators of asymmetric central nervous motor Trunk convexity 1 - 6 points
Cervical rotation deficit 1 - 6 points
control and the results of the eye movement examination
Total score 4 - 24 points
were included in the analysis. For the video documentation,
the infant, together with a physiotherapist, was placed symmetric - asymmetric
supine and prone on a prewarmed mattress with the video Fig. 2 Composition of the asymmetry scale for infants at a post-
camera positioned above them at a distance of 2 m (Fig. 1). term age of 6–16 weeks
160
Fig. 3 a Definition of the six a
categories of trunk convexity.
Categories Spine Pictograms
b Definition of the six categories
of cervical rotation deficits 1 Point No convexity or equal convexity of the spine. or

2 Points Slightly differing convexity of the spine.

3 points Clearly differing convexity of the spine, resolution possible.

4 points Convexity of the spine can be resolved to a straight line.

5 Points Convexity of the spine can be resolved to a flat curve.

6 Points Convexity of the spine can not be resolved.

b
Categories Rotation pictograms

1 Point Free rotation

2 Points Slight head rotation deficit, with a slight resistance during rotation.

3 Points Clear head rotation deficit, preferential head position.

4 Points Restricted head rotation (inner arrow), which may be

intermittently overcome, working area = external arrow.

5 Points Restricted head rotation (inner arrow), which may be intermit-

tently overcome (dotted arrow), working area = external arrow.

6 Points Restricted head rotation (inner arrow), which may

barely be overcome, working area = external arrow.

qualitative analysis of the cervical rotation and cervical tionships among the asymmetry items were evaluated by
lateral flexion was performed by consent of the observers. the t-test.

Pattern definition A predominantly “scoliosis” or the

“torticollis” pattern was assigned if the difference between Results
the sum of trunk convexity prone and supine minus the sum
of cervical rotation deficit prone and supine was ≥3 points. The mean of the three raters of cervical rotation deficit and
For assignment to the “mixed prone” pattern, the difference trunk convexity in the prone and supine position are
between the sum of both prone items and the sum of both presented in Fig. 4. Six infants showed a “scoliosis” pattern,
supine items had to be ≥3 points. The 3-point level was nine infants showed a “torticollis” pattern, 13 infants
chosen by consent of the three observers to be the minimal showed a “mixed prone” pattern and 26 infants showed a
difference which led to the clinical impression that either a “mixed” pattern (Fig. 4). The intraclass correlation coeffi-
torticollis pattern or a scoliosis pattern was clinically leading. cient and the 95% confidence interval of the items are: 0.83
(0.76–0.90) for cervical rotation deficit supine, 0.68 (0.56–
Statistics The inter-observer-reliability of the asymmetry 0.79) for trunk convexity supine, 0.78 (0.70–0.87) for
items was assessed using intraclass correlation. The rela-
 161

Fig. 4 Patterns of postural asymmetry: mean score results of three independent observers of trunk convexity and cervical rotation deficit in
the prone and supine position

cervical rotation deficit prone and 0.79 (0.71–0.87) for right and left occipital bone (14/13 infants, respectively)
trunk convexity prone. was equally affected. Four of these infants showed a stra-
Analysing cervical rotation, a characteristic pattern in the bism which resolved spontaneously at the age of 12 months.
prone and the supine position, was identified in most of the
infants. In the supine position in 49 of 54 infants (91%),
head rotation was always combined with a cervical lateral
flexion to the contralateral side (Fig. 5a). In the prone
position in 36 of 54 (67%) infants, head rotation was always
combined with a cervical lateral flexion to the ipsilateral
side (Fig. 5b). The same pattern was seen in all eight infants
from the first study collective without significant asymme-
try (asymmetry score <12 points), indicating that this
pattern is the physiological cervical rotation pattern in
infants aged 6 to 12–16 weeks. In the prone position, 18
infants (see single asterisk in Fig. 4) showed a fixed cervical
lateral flexion to the contralateral side during a head turn to
either side (Fig. 6a). Only five of these infants had a fixed
lateral flexion to the same direction in the supine position
(n=3 right side, n=2 left side) (Fig. 6b), these infants are
marked with a double asterisk in Fig. 4. In all 18 infants
with a fixed cervical lateral flexion in the prone position, the
side of lateral flexion corresponded to the intrauterine
position. A right-sided cervical lateral flexion was present
in infants with a right occiput anterior position in utero, and
a left-sided cervical lateral flexion was present in infants
with a left occiput anterior position. In total, 24/49 (49%)
infants with a physiological rotation pattern in the supine
position and 24/36 infants (72%) in the prone position
showed a side difference in cervical lateral flexion. There
was no correlation between cervical lateral flexion and
cervical rotation. Fifteen infants (28%) were able to
decrease their cervical rotation deficit at the expense of
the degree of convexity (Fig. 7).
Twenty-seven infants (50%) had a deformational pla- Fig. 5 Physiological pattern of head turning in the first months of
giocephaly, seven of them from the generalized type. The life
162

showed a unilateral hip dysplasia Type II c-g according to

Graf [13].
Evaluation of the side preference of the asymmetric
features showed a side agreement of trunk convexity in the
prone and supine position in 29 of the 54 infants (54%) and
a side agreement of cervical rotation deficit in both posi-
tions in 25 of the 54 infants (46%). Two-thirds of these
congruent infants showed a left-sided convexity (18 in-
fants) and/or a left-sided cervical rotation deficit (17
infants). No further association between side preference
of trunk convexity, cervical lateral flexion and cervical
rotation deficit in either the prone or in the supine position
was detected.
In 19 infants, the right hand reflex, and in five infants the
left hand grasp reflex was unilaterally more intensive. In
50% of these infants (n=12), a concomitant increase of the
foot grasp reflex was observed on the same or opposite side.
Three infants showed a side different foot grasp reflex with
a bilateral symmetric hand grasp reflex. An increased foot
grasp reflex on the left side correlated with a left-sided
cervical rotation deficit in the prone position (p=0.013).

Discussion

Our data show that infantile postural asymmetry quite

commonly has multiple functional and morphological
manifestations which may be focally accentuated. These
manifestations have to be considered in the planning and
Fig. 6 Fixed cervical lateral flexion to the left side in the prone and
evaluation of diagnostic procedures and therapy. The
supine position management of infantile asymmetry should be kept in the
hands of the paediatrician who initiates conservative treat-
ment and refers the infant to specialists for the treatment of
An oblique body position was present in 13 of 54 infants focally aggravated symptoms. If started early, conservative
(24%), and the side of preferential trunk rotation corre- approaches that involve putting the infant in the prone
sponded with the side of the affected occipital bone. An position while awake, feeding from both sides and handling
unilateral calcaneus foot position was registered in 11 have been shown to be promising [1, 2–5, 11, 16, 17, 19,
infants, and an unilateral adductus foot position was reg- 34]. These are useful approaches to forestall the develop-
istered in one infant. In six of these infants an oblique body ment of plagiocephaly and ensuing temporomandibular
position was present with a predominantly left calcaneus joint displacement [31], strabism following frontal plagio-
foot position and right-sided oblique body position pattern. cephaly with desagittalization of the tendon of the superior
Asymmetric foot position and oblique body position were oblique eye muscle [7, 10], torticollis-related head immo-
positively related (p=0.017). Of 54 infants, four (7%) bility [4, 5, 9, 10] followed by asymmetric displacement of

Fig. 7 Reduction of cervical

rotation deficit at the expense of
an increase in trunk convexity
 163

the occipital condyles and atlanto-occipital displacement hand and foot grasp reflexes found in some of our infants
[18] and subsequent progressive scoliosis [4, 9, 21, 23, 32], may be a hint for subtle neurological changes in these in-
foot malpositions with gait disturbances resulting from fants. Follow-up studies of children with infantile postural
oblique body position and hip asymmetry [3, 25, 36]. These asymmetry will be of interest in view of the claim that
complications bear witness to the mutual dependence of, infantile position preferences influence the development of
and interaction between, functional and morphological perceptual and motor preferences by increasing visual
features characterizing infantile postural asymmetry. The orientation to the right side [14, 20, 24, 28, 30].
complexity is further accentuated by changes in body
position with time. As the sidedness of cervical rotation
deficit and trunk convexity differs depending on the prone References
and supine position, postural changes will alter the degree
of cervical rotation deficit and, consequently, trunk con- 1. Argenta LC, Davis LR, Wilson JA, Bell WO (1996) An
vexity. This, in turn, will lead to changes in the asymmetric increase in infant cranial deformity with supine sleeping
position. J Craniofacal Surg 7:5–11
foot position. Changes such as these may explain that the 2. Aufschnaiter D von (1999) Entwicklung und Entwicklungs-
inter-observer reliability was higher for the entire asym- diagnostik. In: Hartmannsgruber R, Wenzel D (eds) Physiother-
metry score than for each single item [27]. Another illus- apie Pädiatrie Band 12. Georg Thieme Verlag, Stuttgart, pp 5–
trative example for the mutual intensification between the 25
3. Behrman RE, Kliegman RM, Jenson HB (2000) Scoliosis,
individual asymmetric features is the clinical course of craniosynostosis, torticollis. In: Nelson textbook of pediatrics,
congenital muscular torticollis. After birth, the muscular 16th edn. W.B. Saunders Co., Philadelphia, pp 1812–1813, pp
torticollis with cervical lateral flexion to the side of the 2082–2086, pp 2089–2091
muscular damage and a head rotation to the contralateral 4. Binder H, Eng GD, Gaiser JF, Koch B (1987) Congenital
muscular torticollis: results of conservative management with
side appears as a single asymmetric symptom. Only a few long-term follow-up in 85 cases. Arch Phys Med Rehabil 68:
weeks later, a deformational plagiocephalus develops 222–225
which is, after verticalization, followed by a double curved 5. Boere-Boonekamp MM, Linden-Kuiper AT van der (2001)
scoliosis with a thoracic convexity on the affected side in Positional preference: prevalence in infants and follow-up after
untreated cases [2, 10]. two years. Pediatrics 107:339–343
6. Brett EM (1991) Paediatric neurology. 2nd edn. Churchill
The left side predominance of trunk convexity and Livingstone, New York
cervical rotation deficit in supine and prone congruent 7. Brodsky MC, Baker RS, Hamed LM (1996) Torticollis and
infants in our study group are in accordance with other head nodding. In: Pediatric neurophthalmology. Springer,
reports [5, 21, 29]. However, a comparison is limited by a Berlin Heidelberg New York, pp 351–368
8. Brunetaeu RJ, Mulliken JB (1992) Frontal plagiocephaly:
non-standardized clinical evaluation of the scoliosis and synostotic, compensational, or deformational. Plast Recontr
positional preference in these reports. The great pattern Surg 89:21–32
variation in the prone and supine position in our infants 9. Canale ST, Griffin DW, Hubbard CN (1982) Congenital
suggests a more complex regulation of postural asymmetry muscular torticollis. A long-term follow-up. J Bone Joint
Surg Am 64:810–816
than previously thought. 10. Cheng JC, Au AW (1994) Infantile torticollis: a review of 624
Our qualitative analyses of cervical rotation revealed that cases. J Pediatr Orthop 14:802–808
during the first 4 months of life, head turn is physiologically 11. Davis BE, Moon RY, Sachs HC, Ottolini MC (1998) Effects of
associated with a contralateral cervical lateral flexion in the sleep position on infant motor development. Pediatrics 102:
supine position and an ipsilateral cervical lateral flexion in 1135–1140
12. Flehmig I (2001) Normale Entwicklung des Säuglings und ihre
the prone position. A cervical rotation deficit in the supine Abweichungen. Georg Thieme Verlag, Stuttgart
position thus points to a sternocleidomastoid and trapezial 13. Graf R (1984) Classification of hip joint dysplasia by means of
muscle dysfunction and/or a fixation in the C1/C2-joint, sonography. Arch Orthop Trauma Surg 102:248–255
whereas a cervical rotation deficit in the prone position 14. Gunturkun O (2003) Adult persistence of head-turning asym-
metry. Nature 421:711
indicates a dysfunction of the autochthonous anterior and 15. Hamanishi C, Tanaka S (1994) Turned head-adducted hip-
posterior cervical muscles and/or fixation in the C2–C7- truncal curvature syndrome. Arch Dis Child 70:515–519
joints [18, 26]. According to our data, cervical rotation 16. Hunt CE, Puczynski MS (1996) Does supine sleeping cause
deficit and asymmetric cervical lateral flexion seem to be asymmetric heads? Pediatrics 97:127–129
17. Kane AA Mitchell LE, Craven KP, Marsh JL (1996) Ob-
independent of each other and, for the purpose of thera- servations on a recent increase in plagiocephaly without synstosis.
peutic goal definition, have to be documented separately. Pediatrics 97:877–885
Cervical rotation deficit can be quantified using the asym- 18. Kapandji IA (1992) Funktionelle Anatomie der Gelenke. Enke,
metry scale, whereas cervical lateral flexion can only be Stutgart
documented qualitatively at that same age group. 19. Kattwinkel J, Brooks J, Keenan MJ, Malloy M (1996)
American Academy of Pediatrics; positioning and sudden
Our observation that the most severe examples of asym- infant death syndrome (SIDS): update. Pediatrics 98:1216–
metric cervical lateral flexion were correlated with the 1218
infantile intrauterine position supports the pathogenic 20. Konishi Y, Mikawa H, Suzuki J (2002) Asymmetrical head-
concept of infantile postural asymmetry as a sequel of turning of preterm infants: some effects on later postural and
functional lateralities. Dev Psychobiol 40:1–13
intrauterine constraint. 21. Lloyds-Roberts GC, Pilcher MF (1965) Structural idiopathic
Although, in general, infants with postural asymmetry scoliosis in infancy: a study of the natural history of 100
are considered to be neurologically healthy, the asymmetric patients. J Bone Joint Surg Am 47-B:520–523
164
22. Mau H (1968) Ist die sogenannte Säuglingsskoliose behan- 30. Skiba M, Diekamp B, Gunturkun O (2002) Embryonic light
dlungsbedürftig? Dtsch Med Wochenschr 43:2051–2053 stimulation induces different asymmetries in visuoperceptual
23. McMaster MJ (1983) Infantile idiopathic scoliosis: can it be and visuomotor pathways of pigeons. Behav Brain Res 134:
prevented? J Bone Joint Surg Am 65-B:612–617 149–156
24. Michel GF, Harkins DA (2001) Postural and lateral asymme- 31. St John D, Mulliken JB, Kaban LB, Padwa BL (2002)
tries in the ontogeny of handedness during infancy. Dev Antropometric analysis of mandibular asymmetry in infants
Neuropsychol 19:41–51 with deformational posterior plagiocephaly. J Oral Maxillofac
25. Niethard FU, Carstens C, Döderlein L, Peschgens T (1997) Surg 60:873–877
Kinderorthopädie. Georg Thieme Verlag, Stuttgart 32. Thompson SK (1980) Prognosis in infantile idiopathic scoli-
26. Pang D, Veetai L. (2004) Atlantoaxial rotatory fixation: part 1 – osis. J Bone Joint Surg Am 62-B:151–154
biomechanics of normal rotation at the atlantoaxial joint in 33. Tunnessen W (1988) Asymmetry. In: Signs and symptoms in
children. Neurosurgery 55:614–626 pediatrics 2nd edn JB Lippingcott, Philadelphia, pp. 509
27. Philippi H, Faldum A, Bergmann H, Jung T, Pabst B, 34. Visscher F, Graaf T van der, Spaans M, Lingen RA van, Fetter
Schleupen A (2004) Idiopathic infantile asymmetry, proposal WPF (1998) Prone position favorable for motor development of
of a measurement scale. Early Hum Dev 80:79–90 infants. Ned Tijdschr Geneeskd 142:2201–2205
28. Reiss M, Reiss G (1997) Lateral preferences in a German 35. Vojta V (2000) Die zerebralen Bewegungstörungen im
population. Percept Mot Skills 85:569–574 Säuglingsalter. Hippokrates, Stuttgart
29. Rosegger H, Steinwendner G (1992) Transverse fetal position 36. Walsh JJ, Morrissy RT (1998) Torticollis and hip dislocation.
syndrome – a combination of congenital skeletal deformities in J Pediatr Orthop 18:219–221
the newborn infant. Paediatr Paediol 27:125–127