The Development of Anterior-Posterior Spinal Curvature in Children Aged 7-12 Years
The Development of Anterior-Posterior Spinal Curvature in Children Aged 7-12 Years
The Development of Anterior-Posterior Spinal Curvature in Children Aged 7-12 Years
DOI: 10.1515/bhk-2016-0011
Summary
Study aim: Body posture changes depend on numerous factors. The purpose of this study was to assess the changes in posture
in the sagittal plane of children aged 7–12 years.
Material and methods: Postural changes were determined by assessing the postural parameters in the sagittal plane. The meas-
urements were recorded in children attending selected primary schools. In total, 562 subjects were included in the study,
including 278 boys and 284 girls. The body posture was assessed photogrammetrically based on the moiré effect, whereby
5 postural parameters were integrated within the measurement methodology.
Results: The results indicated a naturalistic tendency in postural development – the postural parameters changed with age in
a dynamic fashion; and the postural measurements for boys and girls differed in terms of their range.
Conclusions: The study findings can be used for the accurate mapping of normal postural development and, inter alia, for the
detection and diagnosis of an emerging abnormal posture. Importantly, a correct postural assessment algorithm for children
should include the subject’s age and sex.
Author’s address Aleksandra Truszczyńska, Faculty of Rehabilitation, Józef Piłsudski University of Physical Education,
Marymoncka St. 34, 00-968 Warsaw, Poland aleksandra.rapala@wp.pl
Unauthenticated
Download Date | 5/22/17 2:54 PM
Spinal curvature in children 73
were stated: Does spinal curvature depend on age in this commonly referred to as a moiré pattern, is obtained
adolescent group? And are there significant differences re- (Fig.1). The shape of the topographic lines of the graph
garding gender? depend on the shape of the projected area (usually the
patient’s back), and the distance between the grid screen
Materials and methods and the patient. The height of each contour line can then
be determined by assessing the distance between the light
source, the patient and the recording camera. The moiré
In total, 562 subjects were enrolled in the study, in- method allows for the generation of a three-dimensional
cluding 278 boys (49.38%) and 284 girls (50.62%) with image of the subject’s back. This generated image can
a mean age 9.67 ± 1.7, body weight 32.33 ± 6.6 kg and be used for the calculation and analysis of 93 postural
body height 138 ± 10.3 cm. In all cases, the parents/guard- parameters. Such a methodology is non-invasive in its
ians were required to provide a written informed consent nature and offers a relatively efficient and accurate op-
prior to the inclusion of a subject within the study. tion for assessing the body posture in three dimensions.
Inclusion criteria: Healthy children who perform regu- In addition, the computer-aided technique allows for the
lar daily activities and participate in sport only in physical storage and analysis of large volumes of recorded data.
education class. This method is widely used in many countries for the
Exclusion criteria: Obesity, orthopaedic (feet, knee, assessment, gauging treatment procedures and analysing
hips or spine) malformations or injuries, neurological the progression of scoliosis in children [15]. Importantly,
disorders, and participation in any professional sport the method does not replace the currently used x-ray im-
training. aging techniques, but has been reported as an invaluable
The demographic distribution of the cohort group was screening tool.
as follows: 7 year olds – 83 children (14.77%); 8 year
olds – 77 children (13.70%); 9 year olds – 100 children Test operating procedure
(17.79%); 10 year olds – 86 children (15.30%), which Initially, specific anthropometric points were marked
represented the smallest subgroup; 11 year olds 112 chil- on the subject’s back. Such points included the scapular
dren (19.93%), which constituted the largest proportional lower angles, the spinous processes and the posterior su-
representation of the cohort population; and 12 year old perior iliac spines. The patient’s related demographic and
children – 104 subjects (constituting 18.51% of the study medical data was entered into a centralised database. In
group population). The subjects were classified into the re-
spective age groups based on the following principle: e.g. Table 1. Sex distribution in age groups
children whose calendar age at the time of testing was be-
tween 6.5–7.5 years were classified as belonging to the 7 Age number %
year old age subgroup. The distribution regarding gender Girls 32 38.6
in every age group is presented in Table 1. Boys 7 51 61.4
The consent of the Ethical Board at the University of
Sum 83 0.0
Rzeszów was obtained prior to undertaking the current
study. According to the assumptions and purpose of the Girls 42 54.5
research, the exclusion criteria included the following: Boys 8 35 45.5
age of less than 7 or more than 12 years; a lack of written Sum 77 0.0
informed consent; an incomplete medical history or asso- Girls 53 53.0
ciated documentation; and a history of locomotor abnor-
Boys 9 47 47.0
malities.
Sum 100 0.0
Measurement system Girls 44 51.2
In each individual subject, 56 postural measurements Boys 10 42 48.8
were recorded using the photogrammetric method based Sum 86 0.0
on the moiré effect. All of the measurements were con-
Girls 54 48.2
ducted according to commonly accepted principles and
guidelines [17]. The moiré method is based on the prin- Boys 11 58 51.8
ciple of light beam refraction between the grid screen Sum 112 0.0
and its generated shadow, which is projected onto the Girls 59 56.7
patient who is positioned behind the screen. Light in- Boys 12 45 43.3
terference occurs following the propagation of light
Sum 104 0.0
waves through the raster. As a result, a contour graph,
Unauthenticated
Download Date | 5/22/17 2:54 PM
74 J. Drzał-Grabiec et al.
addition, other data relating to the study subject was re- Statistical analysis
corded. This included information relating to the follow- The parameters were compared between the different
ing: the knees and heels; thorax; history of trauma, surgery subgroups. In order to compare the parameters between
and locomotor abnormalities; as well as the gait observed the two gender groups, the non-parametric Mann-Whitney
during clinical observations. U test for independent variables was used. In order to
The patient was then positioned at the appropriate compare the parameters between the age subgroups, the
distance with the subject’s back towards the camera. The non-parametric multiple comparison Kruskal-Wallis test
lights in the room were then switched off. The alignment for independent variables was used. In all cases, a p-value
of the patient was corrected as necessary, in order to en- <0.05 indicated a statistical significance.
sure that the pelvic rotation angle was 0 degrees. The To verify the normality of distribution, the W Shapiro-
generated digital image of the back was then recorded Wilka test was used.
and saved within a designated centralised database. From
the series of images that were generated, an optimal im- Results
age was selected which correlated with the minimal pel-
vic rotation and corresponded with the modal patient’s
posture. Five postural parameters characterising the shape of the
Following the image selection and recording, the ex- anterior-posterior curvatures were analysed. In Tables 2–6,
aminer marked the necessary anthropometric points using the spinal angles divided by the subjects’ age are presented.
specially designated software. A 3-D image was then gen- Significantly significant differences were noted be-
erated. Based on the generated photograms, it was possible tween boys and girls in the 7 year old and 9 year old
to determine several dozen postural parameters. For the groups. The girls had a higher Alpha angle and their lordo-
purposes of the current research study, 5 parameters were sis was bigger (Table 2). The Beta angle was significantly
assessed (see Section 2.4). Fig. 1 presents the method for different between boys and girls only in the 10 year old
the determination of the 5 parameters. group of children i.e. among the boys it was higher (Ta-
Five parameters were included in the analysis, name- ble 3). There were no differences in the Gamma angle be-
ly: Alpha (α) angle – the inclination of the lumbosacral tween the sexes until the age of 10, where significant dif-
vertebral segment; Beta (β) angle – the inclination of the ferences were noted (in all of the age groups from 10–12),
thoracolumbar vertebral segment; Gamma (γ) angle – with the boys having a significantly greater Gamma angle
the inclination of the upper thoracic vertebral segment; (Table 4). The angle of thoracic kyphosis was significantly
the angle of thoracic kyphosis (ATK); and the angle of greater among the girls in 9, 10 and 12 age groups (Ta-
lumbar lordosis (ALL). The angle of thoracic kypho- ble 5). Similar differences were noted for the lumbar lor-
sis was calculated according to the following equation: dosis angle, where there were no differences until the age
ATK = 180 – (β – γ); whereas the ALL was calculated as of 11, after which a significantly greater angle was found
follows: ALL = 180 – (α + β). among the boys (Table 6).
a) b)
C7
KP
PL
LL
S1
Fig. 1. Tested parameters and their derivation (a), and a generated contour map of the back (b)
Unauthenticated
Download Date | 5/22/17 2:54 PM
Spinal curvature in children 75
Unauthenticated
Download Date | 5/22/17 2:54 PM
76 J. Drzał-Grabiec et al.
Descriptive statistics
Parameter
n Mean ± SD Median Minimum Maximum CVa
α angle [deg] 562 13.29 ± 5.21 13.45 –3.1 27 39.18
β angle [deg] 562 8.69 ± 3.74 8.5 –3 25.1 43.06
γ angle [deg] 562 12.88 ± 4.03 12.85 2.2 25 31.3
Angle of thoracic kyphosis [deg] 562 158.32 ± 4.96 158.05 141.4 172.4 3.13
Angle of lumbar lordosis [deg] 562 157.90 ± 5.92 158 135 175.8 3.75
Unauthenticated
Download Date | 5/22/17 2:54 PM
Spinal curvature in children 77
18
16
14
Alpha angle [degrees]
12
10
6
7 8 9 10 11 12
Age [years]
Fig. 2. Alpha angle measurements over time for boys (n = 278) and girls (n = 284)
11.5
11.0
10.5
10.0
Beta angle [degrees]
9.5
9.0
8.5
8.0
7.5
7.0
7 8 9 10 11 12
Age [years]
Fig. 3. Beta angle measurements over time for boys (n = 278) and girls (n = 284)
The basic statistics describing the measurements for there was a slight and statistically insignificant drop in
all of the five parameters are shown in Table 7. For the the value of the Alpha angle between the 11th and 12th
study population, the Alpha angle increased with an in- year of life. With regards to the thoracolumbar vertebral
crease in age. The greatest such increase occurred be- segment, this parameter appeared to decrease with age
tween the 9th and 10th year of life (Fig. 2). Additionally, (Fig. 3). However, the differences between the individual
Unauthenticated
Download Date | 5/22/17 2:54 PM
78 J. Drzał-Grabiec et al.
16
15
13
12
11
10
9
7 8 9 10 11 12
Age [years]
Fig. 4. Gamma angle measurements over time for boys (n = 278) and girls (n = 284)
11.5
11.0
10.5
10.0
KKP [degrees]
9.5
9.0
8.5
8.0
7.5
7.0
7 8 9 10 11 12
Age [years]
Fig. 5. Angle of thoracic kyphosis measurements over time for boys (n = 278) and girls (n = 284)
age groups were generally not statistically significant. an observed reduction occurring between the 10th and
Similar dynamics were observed with the Gamma angle 11th years of life (Fig. 4). With regard to the measure-
parameter measurements, which appeared to increase ments relating to the ATK, the changes in the angle size
with age in the study population, with the exception of appeared to be irregular over time (Fig. 5). With regards
Unauthenticated
Download Date | 5/22/17 2:54 PM
Spinal curvature in children 79
164
162
162
161
160
KLL [degrees]
159
158
157
156
155
154
7 8 9 10 11 12
Age [years]
Fig. 6. Angle of lumbar lordosis measurements over time for boys (n = 278) and girls (n = 284)
Table 8. Comparison of the tested parameters between the gender subgroups in the current study
to the ALL measurements, the size of this parameter larger inclination of the lumber segment (Alpha angle)
appeared to decrease with an increase in age, with the and the angle of thoracic kyphosis, whereas the Gamma
greatest decrease occurring between the 9th and 10th year angle and the angle of lumbar lordosis are significantly
of life (Fig. 6). Thereafter, the ALL increased between smaller.
the 11th and 12th year of life.
When analysing the parameters, statistically signifi- Discussion
cant differences were shown between the subgroups of
boys and girls for 4 out of the 5 tested parameters (Ta-
ble 8). The only exception to this was the Beta angle Postural development in ontogenesis is a dynamic
(the inclination of the thoracolumbar vertebral segment), process. As a result, it can be difficult to assess and
which showed no statistically significant difference be- provide standard values for what constitutes a “nor-
tween the two gender subgroups. The largest statistically mal” posture or a postural abnormality. In the case of
significant difference between the two gender subgroups the frontal plane, postural symmetry is evaluated. How-
was observed with respect to the recorded Gamma and ever, depending on the researcher and the methodology
ATK angles. This may be because the vertebral col- used, it may be arguable what degree of asymmetry in
umn of a female child is characterised by a significantly the paired ossification points should be considered as
Unauthenticated
Download Date | 5/22/17 2:54 PM
80 J. Drzał-Grabiec et al.
significant, borderline or normal. It is even more diffi- observed between 10 and 11 years of age. The ATK an-
cult to determine the normal range of values for the sag- gle reduced and increased in an alternating fashion, with
ittal plane. It is only known that curvatures should have no clearly discernible pattern of development. However,
a mild contour and should be relatively balanced. If any the observed differences were not statistically significant.
normal values have been determined, they have been Conversely, the ALL angle reduced significantly with age,
applied only to a given study population group. Due to with the largest decrease observed between the 9th and 10th
these difficulties, assessments are conducted not with years of life. According to Lang-Tapia et al., men tend to
regard to the specific angular value of the curvatures in exhibit smaller lumbar lordosis angles and larger thoracic
the study group, but rather to the differences in these kyphosis angles when compared to women [11, 12]. This
values between the subgroups which are characterised observation is similar to the conclusions reached by Whit-
by sex, age or level of physical activity. This type of come et al., who explained the size of the lumbar lordosis
information sheds light on what factors posture may be in women in terms of the large pelvic space required for
dependent upon, and what issues can significantly affect a foetus during pregnancy [19]. In addition, Cil et al. were
postural changes. unable to show a relationship between the size of the tho-
Age is one of the factors which affect posture. The pos- racic kyphosis angle and the lumbar lordosis angle with
ture of a child differs significantly from the posture of an the gender of children and adolescents aged from 5 to
adult. Moreover, children at various stages of development 20 years [3].
also differ in terms of their posture. Postural differences In this study, the girls tended to have a larger angle of
between the sexes are generally less obvious due to the lumbar lordosis and a smaller angle of thoracic kyphosis
fact that age is often excluded from such an analysis. Such compared to the boys. This is similar to the results ob-
differences are mostly observed in adolescents in the 2nd tained by Poussa et al [18]. They furthermore measured
critical period of posturogenesis [6]. This paper presents the inclination angles of individual vertebral segments and
an analysis of 5 postural parameters in children aged 7 to a larger inclination of the lumbosacral vertebral segment
12 years. In addition, the parameters were analysed ac- was observed in women compared to men. In contrast, the
cording to the gender subclassification and the differences men tended to have a larger inclination of the lumbotho-
were noted. racic vertebral segment. No gender-specific statistically
Lichota studied the inclination of individual verte- significant differences were found in the inclination of the
bral segments with respect to gender and age [13]. More upper thoracic vertebral segment.
specifically, the author went on to show that the posture The results of the current study partially correlated
of boys was characterised by larger Alpha and Gamma with those obtained by the other authors. In terms of age-
angles, compared to those found in girls; whilst any related changes in the inclination of individual vertebral
gender-related differences with respect to the Beta angle segments, the Alpha and Gamma angles were shown to
was minimal. In a follow-up observation after one year, increase in size. On the other hand, the Beta angle was
the authors reported that the inclination of the anterior- shown to progressively decrease with age. The size of the
posterior spinal curvatures, especially the β angle, had angle of the lumbar lordosis also decreased steadily with
increased significantly. These results were confirmed by age. These result were confirmed in the study by Mac-
Xiong et al. [21]. Furthermore, Wilner and Johnson ob- Thiong et al. [14]. In addition, the size of the angle of tho-
served an age-related increase in the angle of thoracic racic kyphosis appeared to show no correlation with the
kyphosis. However, between 10 and 12 years of age, the age of the individual.
ATK tended to decrease [20]. Cil et al. and Giglio and Non-correlations between the results of the current
Volpon examined the postural characteristics of a popu- study and others published in the literature may partially
lation of children and young people aged between 5 and be explained by the fact that the other studies concentrat-
20 years. The scientists showed that the angle of lumbar ed on different age groups. In addition, it may be difficult
lordosis increased with age [3, 8]. Willner and Johnson to compare our results with those of others, since popu-
also examined a group of 1101 children aged between 8 lation groups from different regions of the world were
and 16 years [20]. These authors observed that the small- enrolled in each of the different studies. Thus far, no data
est angle of thoracic kyphosis appeared in children aged has been published which has attempted to identify en-
10–12 years, and there was a statistically significant in- vironmental factors, such as the geographical locality,
crease in the angle of thoracic kyphosis in the 8, 14 and which may significantly affect the size and distribution
16 year old age groups. of the different postural parameters. Some authors have
In the current study, the Alpha and Gamma angles in- already shown that a decrease in the lumbar lordosis an-
creased steadily with age whilst the β angle reduced stead- gle size is positively correlated with the age of an indi-
ily over time. Only a slight reduction in the β angle was vidual [3, 8].
Unauthenticated
Download Date | 5/22/17 2:54 PM
Spinal curvature in children 81
Unauthenticated
Download Date | 5/22/17 2:54 PM
82 J. Drzał-Grabiec et al.
20. Willner S., Johnson B. (1983) Thoracic kyphosis and Fairbank J. (2008) New method of scoliosis deformity
lumbar lordosis during the growth period in children. assessment: ISIS 2 SYSTEM. Stud. Health Technol. In-
Acta Paediatr.,72: 873-878. form., 140: 157-160.
21. Xiong B., Sevastik J., Hedlund R., Sevastik B. (1998)
Sagittal configuration of the spine and growth of the
posterior elements in early scoliosis. J. Orthop. Res., 12: Received 02.03.2016
113-118. Accepted 16.05.2016
22. Zubovic A., Davies N., Berryman F., Pynsent P., Qurai-
shi N., Lavy C., Bowden G., Wilson-Macdonald J., © University of Physical Education, Warsaw, Poland
Unauthenticated
Download Date | 5/22/17 2:54 PM