Performance of
Preschoolers on the
Pediatric Clinical Test of
Sensory Interaction for
Balance
Pamela K. Richardson,
Sarah W. Atwater, Terry K. Crowe,
Jean C. Deitz
Key Words: assessment process, occupational
therapy. tests, by title, Pediatric Clinical Test
of Sensory Interaction for Balance
The purposes of Ihis sludy were to describe Ihe perfonnance of 40 chiLdren aged 4 and 5 years on Ihe Pediatric ClinicaL Test of Sensory Interaction for BaLance (P-CTSIB) and to determine whether age- and
gender-reLated differences were present. The P-CTSIB
measures standing baLance when sensory input is ~ys­
tematicaL(y aLtered. KruskaL- \'(1aLLis one-way ana(yses
of variance by ranks (p :5 ,05) were used for comparisons by age and gender. When the 4-year-oLds were
compared with the 5~year-oLds, Significant duration
differences were found in 4 of the 6 conditions in the
heeL-toe position of the P-CTSIB. The age-reLated differences on the remaining 2 heeL-toe conditions, as weLL
as on Condition 6 of the feet-together position, approached Significance, Gender d~fferences with 4-yearoLds and 5-year-oLds combined were statisticaLLy nonSignificant 'in aLL instances, however, girLs performed
better on 9 of the 12 conditions of the P-CTSIB. The resuLts indicate that the feet-together position can discriminate between chiLdren without baLance defiCits
and chiLdren with baLance deji'cits. The heeL-toe position is dijficuLt for chiLdren aged 4 and 5 years without baLance deJicits and consequent(v has Limited diagnostiC vaLue for this age group,
T
Pamela K. Richardson, "IS, OTR/L, is an Occupational Therapist,
Clinical Training Unit, Child Development and Mental Retardation Center, University of Washington, \X!J-lO, Seattle, Washington 98195,
Sarah W, Atwater, MI'T. PT. is a Lecturer, Division of Physical
Therapy, Department of Rehabilitation Medicine, University of
Washington, Seattle, Washington.
Terry K. Crowe, PhD. OTR/I .. , is Assistant Professor, Division of
Occupational Therapy, Department of Rehabilitation Medicine, lJniversity of Washington, Seattle, Washington,
.Jean C. Deitz, I'hD. OTR.1, 1''\0'1'\, is Associate Professor, Division
of Occupational Therapy, Department of Rehabilitation Medicine, University of Washington, Seattle, Washington.
This arlie/e was accepledJor publication April 8, 7992,
he abili(y (0 maintain postural s(abili(y, or balance,
is a fundamental and au(Oma(ic process (ha( is basic (0 all funcrional movement. Occupa(ional (herapis(s who work wi(h young children frequently assess and
(rea( defici(~: in pos(ural stabili(y (ha( interfere wi(h (he
child's abili(y (0 participa(e in age-appropria(e activi(ies.
Several pedia(ric mo(Or assessmem rools conrain i(ems
(ha( measure balance in quantifiable (erms. Some of (he
rools mos( frequently used by occupational (herapis(s include (he Peabody Developmemal Moror Scales (Folio &
Fewell, 1983); (he Miller Assessmem for Preschoolers
(Miller, 1982); (he Bruininks-Osere(skyTes( ofMo(Or Proficiency (Bruininks, 1978); and (he Sensory In(egra(ion
and Praxis Tes(s (Ayres, 1987). One-fom balance is assessed in each of (hese rools. O(her balance-related i(ems
include onc:-fom hopping, tandem walking, and distance
jumping, These (es(s measure age-rela(ed moror skill development and permi( comparison wi(h established
norms for skill areas or overall mmor proficiency or both,
Only (he Sensory Integra(ion and Pra...xis Tes(s a((emp( (0
assess (he ma(uri(y of (he ves(ibular, visual, and somawsensory sys(ems and (he quality of sensory imegra(ion.
None of (hese rooJs assess specific sensory contributions
ro (he balance response or moror s(ra(egies (hat affect
balance and movement. Occupa(ional (herapis(s often
supplement standardized (es(s wi(h self-designed clinical
observa(ions, which may include placing (he child on an
unstable surface and ohserving (he response ro penurba(ion of (he suppon surface. Because of (he lack of standardized (es(ing procedures and norma(ive data, (he oc-
'Ihe American journal oj Occupaliol1al (hemp.\'
Downloaded From: http://ajot.aota.org/ on 06/19/2017 Terms of Use: http://AOTA.org/terms
793
cupational therapist's experience and skill in eliciting,
observing, and interpreting balance responses become
critical factors in the assessment of postural stability.
These observations may nOt provide enough information
about sensory or moror stratc8Y selection) particularly
when a child demonstrates more subtle deficits in motor
coordination and balance.
An assessment tool that provides information about
sensory contributions to balance could greatly improve
the occupational therapist's ability to assess and treat
children with balance deficits. However, more information is needed about how children without balance deficits use sensory inputs to select available motor strategies
when placed in sensory conflict situations.
Literature Review
Nashner (1982) described two components of the postural reaction. The first component, sensory organization,
was defined by Nashner as "those processes which determine the timing, direction, and amplitude of corrective
postural actions based upon the convergence of orientation information from visual, vestibular and somatosensory inputs" (p. 358). The second component, muscle
coordination, referred to "processes which determine
the temporal sequencing and the distribution of contractile activity among the muscles of the legs and trunk
which generate supportive reactions" (p. 358). Nashner
and Woollacott (1979) described computerized posturography and electromyography (EMG) studies that evaluated selection of sensory inputs. When subjects are
placed on a platform with the option of a movable platform surface (altered somatosensory input) and the option of a movable visual surround (altered visual input),
the relative influence and weighting of support surface,
visual inputs, and vestibular inputs can be systematically
assessed. A computerized record is made of body sway in
different sensory conditions and muscle activity in response to perturbations of balance.
Forssberg and Nashner (1982) used computerized
posturagraphy and EMG techniques to describe an interaction of vestibular, proprioceptive, and visual inputs, the
weighting of which is context-dependent, to trigger balance reactions. They reported that children younger than
age 7 1/2 years do nOt seem to have a systematic method
for weighting the most appropriate sensory system for
the maintenance of balance. Therefore, children younger
than age 7 1/2 years may sway and lose their balance in the
prescnce of conflicting visual and somatosensory inputs.
Using similar laboratory procedures, Shumway-Cook and
Woollacotl (1985) found a stagelike transition from immature to mature postural responses in nondisabled children. The greatest variability in postural responses (and
inability to correctly recognize conflicting sensory information) was present in children aged 4 to 6 years, with
mature postural control emerging between ages 7 and 10
years. Children shifted from a primary dependence on
visual input to a more adultlike dependence on a combination of visual and somatosensory input. Older children
showed more mature timing and strategy selection in
mQtvr rc~p\Jn~c~, The aUlhor~ concluded that the
variability in performance found in the children aged 4 to
6 years was due to their developing more mature moror
strategy coordination as well as the ability to appropriately select from a variety of sensory inputs.
Romero (1990) evaluated trunk flexor muscle
strength in 108 children aged 3 to 6 years without balance
deficits. Significant differences in strength existed for all
groups, with the greatest gain occurring between the
ages of 4 and 5 years. Romero hypothesized that this large
increase in scores might have reflected the children's
improved ability to coordinate the muscular action, rather than structural or biomechanical factors. This improved ability to coordinate the muscular action responsible for postural adjustments corresponds closely to
Shumway-Cook and Woollacott's (1985) findings of the
emergence of more sophisticated motor strategy coordination for postural responses in this age group.
Shumway-Cook and Horak (1986) adapted the computerized posturography procedures used in the above
studies for use in a clinical environment. In their test, the
Clinical Test of Sensory Interactions for Balance (CTSIB),
standing balance is observed under the following six
conditions:
(heir
1.
2.
3.
4.
5.
6.
Normal
Normal
Normal
Altered
Altered
Altered
vision and support surface
support surface, vision eliminated
support surface, altered vision
support surface, normal vision
support surface, vision eliminated
support surface, altered vision.
Duration of stance and quality of movement are measured. Shumway-Cook and Horak (1986) stated that it is
important to know which sense a person depends on
most for sway orientation and how well a person can
adapt to reliance on the various senses in situations of
intersensory conflict.
Crowe, Deitz, Richardson, and Atwater (1990) adapted the CTSIB to assess sensory contributions to standing
balance in children. The pediatric version is called the
Pediatric Clinical Test of Sensory Interaction for Balance
(P-CTSIB). Test procedures were standardized and interrater reliability was examined with 24 children aged 4 to 9
years without balance deficits. The results of this study
indicated that two raters could reliably score sway and
nominal sway categories. Deitz, Richardson, Atwater,
Crowe, and Odiorne (1991) examined the P-CTSIB performance of 109 children aged 6 to 9 years without balance deficits. Two foot positions, feet-together and heeltoe, were used. The investigators found that in general, all
children aged 6 to 9 years could maintain balance in all
sensory conflict situations in the feet-together pOSition.
794
Downloaded From: http://ajot.aota.org/ on 06/19/2017 Terms of Use: http://AOTA.org/terms
Septemher 1992, Volume 46, Numher 9
The heel-toe position resulted in more variability, with
progression of difficulty related to manipulation of visual
cues. No clear developmental progression was found for
the P-CTSIB, although the youngest children scored the
same or lower than the oldest children for duration of
stance in all conditions. Results related to differences in
gender were also inconclusive. The authors recommended further research on larger numbers of children with
and without balance deficits before drawing conclusions
about balance deficits in specific children.
The P-CTSIB can provide specific information on
sensory selection strategies and their relationship to postural responses that has not been readily available to
occupational therapy clinicians. Further descriptive information on the P-CTSIB is needed before it can be used
reliably with preschoolers and kindergartners. The purpose of this study, therefore, was to describe the performance of 4-year-old and 5-year-old children without
balance deficits on the P-CTSIB. The two research hypotheses tested were as follows:
1. There is a significant difference between the performance of 4-year-olds and that of 5-year-oJds on
the P-CTSIB.
2. There is a significant difference between the performance of 4-year-old and 5-year-old girls and
that of 4-year-old and 5-year-old boys on the
P-CTSIB
Method
Subjects
Subjects were 40 preschool and kindergarten children
whose ages ranged from 4 years 0 months to 5 years 11
months. They were recruited from Headstart preschool
programs, a private preschool and day-care facility, a private kindergarten, and a public school kindergarten in
Seattle. Information on parental educational level was
obtained for 38 of the mothers and 33 of the fathers. The
median level for both groups was 2 years of college.
Permission slips describing the testing procedures
were sent home, and the children who received permission to participate were divided into four categories: 4year-old girls, 4-year-old boys, 5-year-oJd girls, and 5-yearold boys. Ten subjects were randomly selected from each
of these groups. The 4-year-old age group contained 13
white children, 6 black children, and 1 Hispanic child; the
5-year-oJd age group, 16 white children and 4 black
children.
AJI parents signed consent forms that were approved
by the University of Washington Human Subjects Committee. Parents indicated by questionnaire responses that
all children were free of major developmental concerns
such as serious problems with motor coordination, seizures, neurological problems, and learning or physical
disabilities. In addition, all children passed a biomechanical screening consisting of strength and range-of-motion
tests of the trunk and lower extremity before the balance
tests were administered.
instrumentation and Procedure
Subjects were barefoot for all testing. The P-CTSIB was
administered as part of a battery of balance tests including
tiltboard tip and one-foot balance. Information regarding
the latter two tests is not reported in this article. The
order of administration of all tests was randomized across
subjects.
The P-CTSIB evaluates duration of standing balance
and amount of body sway under six different sensory
conditions. Combinations of three visual and t,vo support
surface variables are used. The visual variables are (a)
eyes open, normal visual input; (b) eyes closed, visual
input eliminated; and (c) sway-referenced vision, in
which a visual conflict dome moves in phase with the
child's head movement and prevents visual orientation to
the environment. Peripheral vision is restricted at the top,
bottom, and sides, as described by Shumway-Cook and
Horak (1986). A tape mark is placed on the forward part
of the dome as a visual reference point. The support
surface variables are (a) standing on a hard, flat surface:
normal somatosensory input; and (b) standing on a firm,
compliant medium-density foam: inaccurate somatosensory input.
In the first three conditions, the subject stands on a
normal surface with eyes open (Condition 1), with eyes
closed (Condition 2), and with the visual conflict dome
(Condition 3). The three visual conditions are repeated
with the subject standing on the foam (Conditions 4, 5,
and 6). The difficulty of the task is thought to increase
with each condition as sensory information is systematically altered. Conditions 5 and 6 are considered to be the
most difficult because visual and somatosensory input are
eliminated or compromised and thus the child must rely
primarily on the vestibular system to cue a motor response for maintenance of balance.
The six conditions were administered in two positions: (a) feet together, medial malleoli touching; and (b)
heel-toe (the preferred foot is placed behind the non preferred foot with the toes touching the heel). Thus, 12
conditions were tested, The sL,<: conditions for each foot
position were always administered in chronological
order.
Each of the sL,<: conditions was tested twO times in
each of the heel-toe and feet-together positions. The suhject stood with hands on hips, and duration of standing
was measured until the subject had maintained the pOSition for 30 sec or made a postural adjustment. Apostural
adjustment was defined as removing hands from hips,
moving one or both feet from the original positions,
Tbe American ]uumul of Occupational Tberapy
Downloaded From: http://ajot.aota.org/ on 06/19/2017 Terms of Use: http://AOTA.org/terms
795
opening eyes during the eyes-closed conditions, or requiring assistance from the examiner to prevent a fall.
A backdrop with lines radiating in 1° increments
from a central axis at the floor was placed behind the
subject to measure amount of sway, as shown in Figure 1.
A maximum of 40° total sway (20° in each direction) was
measured. Three additional nominal sway scores were
possible: (a) inability of the child to assume the condition
position (i.e., the examiner could not let go of the child);
(b) a fall during the test condition in which the child made
no postural adjustment (a "timber" fall); and (c) inability
of the child to stand in the condition position longer than
3 sec. The latter made it difficult or impossible for the
examiner to determine the degrees of sway that occurred
because balance was maintained so briefly. For the feettogether posi tion of the P-CTSIB, total anterior and posterior degrees of sway were recorded. In the heel-toe position of the P-CTSIB, total lateral degrees of sway were
recorded. Interrater reliability was examined for the sway
measurement in a previous study (Crowe et a!., 1990).
Spearman rank order correlations were used as indexes
of reliability and ranged from .69 for P-CTSIB feettogether Condition 3 to .92 for P-CTSIB heel-toe Conditions 1 and 5
On each of the six P-CTSIB feet-together and six
P-CTSIB heel-toe conditions, the best of the two trials was
recorded. The best trial was defined as the trial with the
longest duration or, if the durations were the same, the
trial with the smallest sway.
Though a quality measure was included in the adult
version of this test, previous research (Crowe et aI., 1990)
suggested that the quality measure was not reliable.
Therefore, quality data are not reported for the children.
Two examiners were required to administer the
P-CTSIB: The ["lfimary examiner recorded sway, and the
secondary examiner pOSitioned the subject, guarded
against falls, and recorded the duration of stance. A digital
stopwatch was used to record time. Other equipment
included a 6-ft backdrop with degree lines (to measure
postural sway), an 18-in by 18-in by 3-in piece of medium
density foam, a visual conflict dome, and a head pointer.
Paper surgical caps were placed on the subjects' heads for
sanitary reasons. The total time required to administer
the test battery was approximately 30 min.
Examiners
The first and second authors served as primary examiners. Both had more than 8 years of clinical pediatric experience and participated in development of the test.
Five people served as secondary examiners. Two
(the third and fourth authors) were registered occupational therapists with pediatric experience and three
were occupational therapy or physical therapy students.
All secondary examiners received training and established
procedural reliability before participating in data collection. In addition, timing of conditions was compared with
that of experienced secondary examiners. Timing agreement had to be within 1 sec.
Results
[lJJ
[15]
1ul,.. ,~.""
Initial alignment
~'"
P-CTSIS
feet together
1•. ,
Alignment for
))"
P-CTSIS
1•. ~
Initial
Alignment and
all testing
P-CTSIS/Heel-Toe
'" 2'lape
feet together
p"''''"
ill
."
j
prelerredlool
Heel-Toe posllion
Figure 1. Alignment for testing and backdrop used for
measuring degrees of sway for the Pediatric Clinical Test of
Sensory Interaction for Balance. Note. From "Interrater Reliability of the Pediatric Clinical Test of Sensory Interaction
for Balance" by T.K. Crowe, J.C. Deitz, PK Richardson, &
S.w. Atwater, 1990, Physical and Occupational Therapy in
Pediatrics, 10(4) p. 10. Copyright 1990 by The Haworth
Press, Inc., 10 Alice St., Binghamton, NY 13904. Adapted
by permission.
Descriptive data were examined first for duration and
second for sway. Because of the skewed score distributions for both duration and sway data, medians and low
and high scores are presented as well as means and standard deviations. Additionally, scores in the 25th percentile are presented for duration to assist in clinical interpretation of scores. For sway, scores in the 75th percentile
are presented, because a high score is more indicative of
dysfunction. Descriptive data for Conditions 5 and 6 for
the feet-together position and Conditions 1 through 6 for
the heel-toe position duration are presented in Table 1.
Data are not presented for Conditions 1 through 4 for the
feet-together position because all but 6 children could
maintain balance for the maximum time for these
conditions.
Descriptive data for degrees of sway are shown in
Table 2. In some conditions the sample sizes are smaller
than those reported for duration because sway was not
measured unless a child maintained balance for more
than 3 sec. For data on the percentages of children (for
each age and condition) who could maintain balance for
more than 3 sec, see Figure 2.
A Kruskal-Wallis one-way analysis of variance by
796
Downloaded From: http://ajot.aota.org/ on 06/19/2017 Terms of Use: http://AOTA.org/terms
September 1992, Volume 46, Numher 9
Table 1
Subjects' Standing Balance Duration on the P-CTSIBa
tween 4-year-olds and 5-year-olds or between boys and
girls,
Duration (in sec)
Agel;
/1'1
Median
SD
Low/25th
perccnrilelHigh
rEI"r-ToGETHER POSITION
Condition 5
4-yr-olds
5-yr-olds
Condition 6
4-yr-olds
5-yr-olds
24
27
30
30
9
7
5/14/30
8/30/30
27
29
30
30
6
4
14/22/30
11/30/30
HeEL-TOE PosrnON
Condition
4-yr-olds
5-vr-olds
Condition
4-yr-olds
5-yr-olds
Condition
4-yr-olds
5-YI'-oJds
Condition
4-yr-olds
:;-vr-olds
Condition
4-yr-olds
5-w-olds
Condition
4-yr-okJs
5-yr-olds
1
017130
18
25
23
30
11
10
1/22/30
6
12
5
10
7
9
0/1/30
1/3/30
5
II
4
9
6
10
0/0/19
0/2/30
12
21
11
23
10
10
0/21.30
0/13/30
4
6
3
7
5
6
2/4/13
0/3/27
1
4
0
2
2
7
0/0/6
0/1124
Heel- Toe Position
This foot position was much more difficult for the subjects than the feet-together position. In Condition 6, only
6 of the 40 subjects were able to balance for more than 3
sec.
Analysis of duration by age (see Table 3) revealed
significant differences on Conditions 1, 2, 4, and 6. Conditions 3 and 5 approached significance (p < .09 and
p < .06, respectively). Analysis of duration by gender did
not reveal any significant differences between boys and
2
3
Table 2
P-CTSIBa Sway in Degrees by Age
4
5
6
No/e. Condition 1: Normal suppon surfacc, normal vision; Condition 2:
Normal suppon surface, vision eliminated; Condition 3: NOI'mal suppon surface, altered vision; Condition 4: AJtered support sUI'face, normal vision; Condition :;: Altered suppon surface. vision eliminated;
Condition 6: Altered suPPOrt sUI'face. altered vision.
"P-CTSlB = Pediatric Clinical Test of SensoJ'\' 11l(eraction for Balance.
= 20 subjccts for each group.
.
"n
ranks (Siegel, 1956) was used to test the research hypotheses relating to age and gender. This nonparametric statistic was used because of the skewed score distributions.
We combined age groups when comparing performance
by gender, and we combined gender groups when comparing performance by age. The alpha level was set at
p :5 .05 (two-tailed).
Fee/-Together Position
In Conditions 1, 3, and 4, all 40 subjects were able to
balance for the fuJJ 30 sec. In Condition 2, three 4-year-old
girls balanced for 4, 8, and 16 sec; twO 4-year-old boys
balanced for 20 and 26 sec; and one 5-year-oleJ girl balanced for 22 sec. The remaining 34 subjects balanced for
30 sec; in Condition 5, 12 of the 40 subjects balanced for
fewer than 30 sec; in Condition 6, 7 of the subjects balanced for fewer than 30 sec. No significant differences in
duration were found for gender or age for all feettogether conditions. However, for Condition 6, the duration difference between 4-year-olds and 5-year-olds approached significance (p < .06). Analysis of sway by age
and by gender revealed no significant differences be-
Agc
,H
Median
SD
Lowl7:;th
Percentilci
High
n
Ft,:r-ToGETIIER PosrlloN
Condition 1
4-year-olds
5-vear-olds
Condition 2
4-vear-olds
5-yca r-olds
Condition 3
4-year-olds
5·year-olds
Condition 4
4'\'ear-olds
5,vear-okb
Condition 5
4-vear-olds
5-vear-olds
Condilion 6
4·year·okb
5,year-olds
3
2
2
2
1
2
1/3/5
0/3/6
20
20
4
3
4
2
I
1/6/9
2/4/8
20
20
1/6/9
2/6/8
20
20
4
3
4
4
4
3
4
,3
2
2
2/5/8
2/4/9
20
20
5
6
4
6
1
3
3/5/8
317117
20
20
7
7
7
7
2
2
3/8/11
4/8/12
20
20
HtEL-ToE
Condition 1
4-vear-olds
5·year-olds
Condition 2
4-vear-olcls
5,year-olcls
Cond ition 3
4, vea r-olds
5-)'eal'-0Ids
Condition 4
4-veal'-0Ids
5-vear'-olds
Condition 5
4-vear-okb
5-ycar-olds
Condition 6
4-\'ear·olds
5-\'ear-olds
POSlTlO~
8
8
7
6
5
5
2/10/17
2/13/19
18
18
10
14
7
9
9
10
2/17/26
2/23/35
11
15
10
8
8
7
8
4
4/14/26
3/13/15
10
12
10
7
7
7
7
3/13/26
111 (J/22
14
19
7
8
5
6
6
2110/13
0/12/23
9
14
11
7
11
6
2
2
9/- h/12
5/8/9
2
4
5
4
No/e. Condition 1: Normal suppor't sur'face, normal visiun; Condition 2:
Nonnal ,uppon sUI'face, vision eliminated; Condition 3: Normal suppon surface. altered vision; Condition 4: AJtered support surface, normal vision: Condllion :;: AJtered suppon surface, vision eliminatcd;
Condition 6: Altercel suppon surface, alrered vision.
"P-CTSlB = Pediatric Clinical Test of Sensory Interaction for Balance,
hNo SCOi'e for the 75th percentile because n = 2
Tbe American Juurnal o/Occupaliunal Tberapr
Downloaded From: http://ajot.aota.org/ on 06/19/2017 Terms of Use: http://AOTA.org/terms
797
90 -
• 4·year olds
BO
o 5·year olds
70
Table 3
Kruskal-Wallis One-Way Analysis of Variance for PCTSlsa Duration by Age
60
% of SUblects 50
balancing > 3
seconds
FEET-TOGETHER POSITION
40
3D
20
10
2
3
5
6
Condition
Figure 2. Percentage of subjects balancing 3 sec. or more
on the Pediatric Clinical Test of Sensory Interaction for Balance in the heel-toe position. Condition 1: Normal vision
and support surface; Condition 2: Normal support surface,
vision eliminated; Condition 3: Normal support surface, altered vision; Condition 4: Altered support surface, normal
vision; Condition 5: Altered support surface, vision eliminated; Condition 6: Altered support surface, altered vision.
Condition 1
4-year-olds
5-year-olds
Condition 2
4-year-olds
5-year-olds
Condition 3
4-year-olds
5-yea r-olds
Condition 4
4-year-olds
5-year-olds
Condition 5
4-year-olds
5-year-olds
Condition 6
4-year·olds
5-year-oJds
205
205
0.00
1.00
9.5
115
2.11
<0.15
205
20.5
000
1.00
205
205
0.00
1.00
18.4
22.6
1.92
<0.17
18.2
22.8
369
<006
HEEL-TOE POSITION
girls, although girls received higher median scores on five
of the six conditions. Analysis of sway revealed no significant differences between 4-year-olds and 5-year-olds, or
between boys and girls.
Discussion
Feet-Together Position
Most of the subjects were able to balance for the maximum amount of time on Conditions 1 through 4, and
there was limited variability between subjects. Conditions
5 and 6 were more difficult, and subjects showed more
variability in performance on these conditions. Conditions 5 and 6 require a primary dependence on the vestibular system for maintaining balance. Although Condition 6, with unreliable visual and somatosensory input, is
assumed to be the most difficult condition of the test, the
subjects in this sample balanced for less time in Condition
5, where vision is absent and somatosensory input is
unreliable. However, the difference was not significant.
ThiS result may suggest that, as Shumway-Cook and
Woollacott (1985) found, 4-year-olds and 5-year-olds depend more on vision to mediate their balance response.
Conditions 2, 5, and 6 were the only conditions that
demonstrated any variability. The two eyes-closed conditions (Conditions 2 and 5) were among the mOSt difficult
for the subjects. Four-year-olds experienced more difficulty with these two conditions than did 5-year-olds,
which may reflect an increased ability of 5-year-olds to
select other sensory inputs to maintain balance. In general, the feet-together position does not discriminate well
between 4-year-olds and 5-year-olds or between boys and
girls, mainly because much of the test can be easily accomplished by most children without balance deficits in
thiS age group.
Condition 1
4·year-olds
5-year-olds
Condition 2
4-year-olds
5·year-olds
Condition 3
4·year-olds
5-year-olds
Condition 4
4-year-ulds
5-year-olds
Condilion 5
4·year-olds
5-year·olds
Cundition 6
4-year-olds
5·year-olds
165
245
5.29
.022
16.2
24.8
5.51
019
173
237
2.98
.084
157
253
690
009
17.0
240
3.64
.057
15.7
253
7.47
006
Note. Chi-square values and significance levels are correCled for ties.
Condition 1: Normal support surface, normal vision; Condilion 2: Normal support surface, vision eliminated; Condition 3: Normal support
surface, allered vision; Condition 4: Altered support surface, normal
vision: Condition 5: Altered support sLlIface, vision eliminated: Cundition 6: Altered support surface, altered vision.
:Ip·CrSIB = Pedialric Clinical Tl:SI of Sensory Interaction for Balance.
hn = 20 for each group.
.
Heel- Toe Position
For each condition of the heel-toe pOSition, certain children were unable to assume and maintain the position for
1 sec or more. From the trends of the median scores (see
Table 1), it appears that the availability of visual cues
affected standing balance duration. The decrease in
scores from Condition 1 to Condition 2 may be related to
the elimination of visual cues in Condition 2. Condition 4
introduces an unreliable support surface; however, with
vision restored in this condition, performance improves
over Condition 3. It appears that as long as normal visual
input is available, children aged 4 and 5 years can generally override unreliable somatosensolY cues to maintain
798
Downloaded From: http://ajot.aota.org/ on 06/19/2017 Terms of Use: http://AOTA.org/terms
September 1992, Volume 46, Number 9
balance. However, as seen in Conditions 5 and 6, when
visual input is absent or conflicting and inaccurate, these
children seem to be unable to disregard the inaccurate
input and select vestibular inputs to maintain balance.
In all conditions, 5-year-olds performed better than
4-year-olds, suggesting an overall maturation effect.
When comparing the performance of children aged 6 to 9
years on the P-CTSIB (Deitz et ai, 1991) with that of the
subjects in this study, we observed that the 5-year-olds
appeared to be more similar to the 6-year-olds than they
were to the 4-year-olds. Proportionately fewer 4-year-olds
and 5-year-olds were able to maintain their balance in the
presence of sensory conflicts when compared with the
older age grou ps. Ninety percent of the children aged 4 or
5 years were able to maintain the heel-toe position for
more than 3 sec in Condition 1, as opposed to all but 1 of
the 109 children aged 6 to 9 years. This finding probably
indicates that assuming the heel-toe position is difficult
for some children in this younger age group. When the
sensory conflict conditions were administered, many
more children aged 4 or 5 years than those aged 6 to 9
years balanced for 3 sec. or fewer or were unable to
assume and maintain the position. This finding appears
consistent with Shumway-Cook and Woollacott's (1985)
finding that a transition in sensory selection strategies
occurs from age 4 years to age 6 years. A few children
appear to have completed the transition at ages 4 years
and 5 years; however, many more children are balancing
proficiently at age 6 years, with some improvement up to
age 9 years. The high degree of variability in scores for 4year-olds and 5-year-olds on Conditions 2 through 6 may
reflect that children in these age groups are at different
stages in this transition and do not yet have a systematic
sensory selection strategy.
The P-CTSIB heel-toe pOSition appears to be difficult
for 4-year-olds and 5-year-olds for several reasons. Reliance on the visual system to mediate the balance response appears to be an important reason. Additionally,
biomechanical and motor control factors must be considered. Children's shorter stature requires more rapid and
frequent corrections of sway motion (Forssberg &
Nashner, 1982) The heel-toe position reqUires the ability
to balance on a very narrow base of support as well as the
fine coordination of ankle and hip strategies. As Black,
Wall, and Nashner (1983) found in adults with vestibular
deficits, use of appropriate motor strategies in an altered
sensory environment depends on the ability to use vestibular input appropriately. Young children who cannot
yet select among competing sensory inputs may be similarly compromised in their ability to select an aplJropriate
motOr strategy. Forssberg and Nashner (1982) noted that
children younger than age 71/2 years do not have a systematic method for weighting the most appropriate sensory
inputs. Instead, young children may randomly change the
weighting of support surface, vestibular, and visual inputs. As a result, performance in balance-related tasks
with children younger than age 71/2 years shows more
variability than with children older than age 7 1/2 years.
Clinical Implications
Most of the children in this sample were able to maintain
balance in the feet-together position under all sensory
conditions and had mean total sway of7° or less. If a child
is unable to balance in sensory conflict situations in the
feet-together pOSition or demonstrates increased body
sway or both, this may indicate that difficulties in sensolY
selection strategies result in inability to coordinate motor
strategies for standing balance. The feet-together position of the P-CTSIB, therefore, can discriminate between
children with age-appropriate balance responses and
those with balance deficits. The heel-toe position was
much more difficult for the 4- and 5-year-olds, as indicated by the 25th percentile scores. With the exception of
Conditions 1 and 4, no 25th percentile score is higher
than 4 sec. Because of the difficulty experienced by the
children without balance deficits on this task, the heel-toe
pOsilion cannot he used diagnostically for children aged 4
and 5 years.
Directions for Future Research
The relationship of sensOlY environment 10 motor strategy selection should be explored, as should the relationship of various body size parameters to standing balance.
A more reliable measurement system for motor strategies
also should be developed, and the method for measuring
sway should be further examined. Further testing of the
P-CTSIB on a larger number of children both with and
without a variety of neuromOlor and sensorimotor deficits is also necessary...
Acknowledgments
We thank the children, rarents, and teachers who participated
in this study as well as the senior occupational therapy and
rhysical thel'apy students who served as secondary examiners.
This study was funded in pan by a grant from the American
Occupational Therapy Foundation and the American Occupational Therapy Association as pan of the Developmental Disahilities Research Symposium.
This paper was hased on a thesis completed by Pamela K.
Richardson in partial fulfillment of the master of science degree
in Rehahilitation Medicine, Division of Occurational Therapy,
University of Washington, Seattle.
References
Ayres, A. j. (1987). Sensor)' Integration and Pmxis Tests.
Los Angeles: Western Psychological Services.
Black, F. 0, Wall. C, & Nashner, L. M. (1983) Effects of
visual and support surface orientation references upon postural
conrrol in vestibular deficient subjects. Acta Olo-Lmyngotogica,
95, 199-210
Bruininks, R. H. (1978). Bruininks-Oserelsky Test o/IVIotor
ProfiCiency. Circle Pines. MN: American Guidance Service.
The Aillerican juurnal uj" Occupatiunal Themp) ,
Downloaded From: http://ajot.aota.org/ on 06/19/2017 Terms of Use: http://AOTA.org/terms
799
Crowe, T. K., Deitz, J. c., Richardson, P" &. Atwatel', S. W
(1990). Interrarer reliability of the Pediatric Clinical Test of Sensory Interacrion for Balance. Phvsical and Occupational Ther-
apy in Pedialrics, 10(4), [-27.
Deirz,.!. c., Richardson, P. R, Arwater, S W, Crowe. T. K"
& Odiorne, M. (1991). Performance of normal chIldren on the
Clinical Test of Sensory Interaction for Balance. Occupational
Tberapv Joumal of Research, 11, 336-3'56.
Folio, M. R, & Fewell, R. R. (1983). Peahodv Del'elopmental Motor Scales and Activity Cards. Allen, TX: Develormental
Learning Materials.
Forssberg, H., &. Nashner, L. M. (1982). Ontogenetic developmenr of rostural control in man: Adaptation to altered surport and visual conditions during stance. Journal of Neuruscience, 2, '545-5'52.
Miller, L J. (1982). Nliller Assessment for Preschoulers.
Littleton, co: Foundation for Knowledge in Development.
Nashner, L. M. (1982). Aclaptation of human movemeI1l
to altered envimnments. Trends In Neurosciences. 5, 3'58-
361
:'-1ashner. L. lvI., &. WooJlacott, M. H. (1979) The organization of capid postural adjustmenrs of standing humans: An experimental conceptual model. In R E. Talbot and D. R Humphre;' (EdS), Posture and mucement (pp. 243-357). New York:
Raven.
Romero, P W. (1990). Trunk flexor muscle strength in
healthv children 3 to 6 yeal-s okL Pediatric Pbvsical Therapl',
2(1), 3-9
Shumway-Cook, A., &. Horak, F. B. (1986). Assessing [he
influence of sensory interaction on balance: Suggestions from
the field. Pbysicul Therapl', 66, 1548-1550.
Shumway-Cook, A., &. Woollacort, M. H. (1985). The
growth of stabiliry: Postural control from a developmental rerspeerive. Journal of Motor Behal'ior, 17, 131-147.
Siegel, S. (1956). Nunparametric statistics/ur the behal'loral sciences. New York: McGraw-Hili.
Coming in October:
• A holistic approach to an occupational adaptation practice
model
• Clinical reasoning process for service provision in public
schools
• Cross-cultural assessment of process skills
• Deficits of reaching in subjects with left hemiparesis
UPDATE
• Continuing education reqUirements to maintain occupational
therapy licensure
Turn to AJOT for the latest information on occupational
therapy treatment modalities, aids and equipment, legal and
social issues, education, and research.
800
Downloaded From: http://ajot.aota.org/ on 06/19/2017 Terms of Use: http://AOTA.org/terms
Seplemher 1')<)1. volullle 46. Numher 9