Research Report

Comparing Two Motor Assessment

Tools to Evaluate Neurobehavioral
Intervention Effects in Infants With
Very Low Birth Weight at 1 Year
Janeline W.P. Van Hus, Martine Jeukens-Visser, Karen Koldewijn,
Loekie Van Sonderen, Joke H. Kok, Frans Nollet, Aleid G. Van Wassenaer-Leemhuis

Background. Infants with very low birth weight (VLBW) are at increased risk for
motor deficits, which may be reduced by early intervention programs. For detection
of motor deficits and to monitor intervention, different assessment tools are available.
It is important to choose tools that are sensitive to evaluate the efficacy of intervention on motor outcome.

Objective. The purpose of this study was to compare the Alberta Infant Motor
Scale (AIMS) and the Psychomotor Developmental Index (PDI) of the Bayley Scales
of Infant DevelopmentDutch Second Edition (BSID-II-NL) in their ability to evaluate
effects of an early intervention, provided by pediatric physical therapists, on motor
development in infants with VLBW at 12 months corrected age (CA).

Design. This was a secondary study in which data collected from a randomized
controlled trial (RCT) were used.

Methods. At 12 months CA, 116 of 176 infants with VLBW participating in an RCT
on the effect of the Infant Behavioral Assessment and Intervention Program were
assessed with both the AIMS and the PDI. Intervention effects on the AIMS and PDI
were compared.

Results. Corrected for baseline differences, significant intervention effects were

found for AIMS and PDI scores. The highest effect size was for the AIMS subscale sit.
A significant reduction of abnormal motor development in the intervention group
was found only with the AIMS.

Limitations. No Dutch norms are available for the AIMS.

Conclusions. The responsiveness of the AIMS to detect intervention effects was
better than that of the PDI. Therefore, caution is recommended in monitoring infants
with VLBW only with the PDI, and the use of both the AIMS and the Bayley Scales of
Infant Development is advised when evaluating intervention effects on motor development at 12 months CA.

J.W.P. Van Hus, PT, Department

of Rehabilitation A-01, Academic
Medical Centre, University of
Amsterdam, Post Box 22660,
1100 DD, Amsterdam, the Netherlands. Address all correspondence to Mrs Van Hus at:
j.w.vanhus@amc.uva.nl.
M. Jeukens-Visser, PhD, Department of Rehabilitation, Academic
Medical Centre, University of
Amsterdam.
K. Koldewijn, PhD, Department of
Rehabilitation, Academic Medical
Centre, University of Amsterdam.
L. Van Sonderen, MD, Department of Neonatology, Emmas
Childrens Hospital, Academic
Medical Centre, University of
Amsterdam.
J.H. Kok, MD, PhD, Department of
Neonatology, Emmas Childrens
Hospital, Academic Medical Centre, University of Amsterdam.
F. Nollet, MD, PhD, Department of Rehabilitation, Academic
Medical Centre, University of
Amsterdam.
A.G. Van Wassenaer-Leemhuis,
MD, PhD, Department of Neonatology, Emmas Childrens Hospital, Academic Medical Centre,
University of Amsterdam.
[Van Hus JWP, Jeukens-Visser M,
Koldewijn K, et al. Comparing two
motor assessment tools to evaluate neurobehavioral intervention
effects in infants with very low
birth weight at 1 year. Phys Ther.
2013;93:14751483.]
2013 American Physical Therapy
Association
Published Ahead of Print:
June 13, 2013
Accepted: June 10, 2013
Submitted: November 19, 2012

Post a Rapid Response to

this article at:
ptjournal.apta.org
November 2013

Volume 93

Number 11

Physical Therapy f

1475

Evaluating Efficacy of Intervention on Motor Outcome

nfants with very low birth weight

(VLBW) have more motor deficits
compared with their full-term
counterparts, and their motor deficits persist throughout childhood.1
Some studies have demonstrated
poor quality of movements, low postural control, and atypical postures
of infants with VLBW in their first
year of life.2,3 Therefore, motor
development is an important domain
to follow up on in these infants. In
2006, the American Academy of
Pediatrics published guidelines for
the follow-up of preterm infants and
recommended that all infants with
VLBW should have a structured, ageappropriate neuromotor examination at least twice during the first
year of life.4 However, no recommendations for specific instruments
were made.
Until recently, there was little evidence of an effect of early intervention programs on motor outcome in
infants with VLBW.5 A systematic
review of neuromotor assessments6

concluded that large-scale randomized controlled trials (RCTs) of interventions are needed, as well as
assessment tools that are sensitive
enough to measure change in motor
performance, in order to evaluate
the efficacy of the intervention programs in the first year of life. The use
of more than one assessment tool is
recommended to ensure appropriate
predictive, discriminative, and evaluative assessments.6
Between 2004 and 2007, a multicenter RCT7 was designed and conducted by pediatric physical therapists to evaluate the effectiveness of
the Infant Behavioral Assessment
and Intervention Program (IBAIP)8
in infants with VLBW. The instrument used to measure the primary
outcome at 6 and 24 months corrected age (CA) was the Bayley
Scales of Infant DevelopmentDutch
Second Edition (BSID-II-NL).9 At both
time points, an intervention effect
was found on the motor domain.7,10

The Bottom Line

What do we already know about this topic?
Infants with very low birth weight (VLBW) are at risk for motor deficits,
deficits that may be reduced by early intervention programs. In order to
evaluate the efficacy of intervention on motor development, it is important to choose assessment tools that are sensitive.

What new information does this study offer?

The authors compared the Alberta Infant Motor Scale (AIMS) and Bayley
Scales of Infant DevelopmentSecond Edition (BSID-2) at 12 months
corrected age and found intervention effects with both assessment tools.
However, a reduction of abnormal motor development was reflected only
on the AIMS, and the responsiveness of the AIMS was better than the
BSID-2.

If youre a caregiver, what might these findings mean

for you?
Parents of infants with VLBW may gain better insight into their motor
development if both the AIMS and BSID-2 assessments are used.

1476

Physical Therapy

Volume 93

Number 11

At 12 months CA, motor outcome

was measured using the Psychomotor Developmental Index (PDI)
of the BSID-II-NL, and in a large subset of infants participating in the
trial, the Alberta Infant Motor Scale
(AIMS) also was used to measure
motor outcome.11 The AIMS was
added to the assessment protocol
because information about possible
overestimation of the PDI12 became
apparent when the 12-month
follow-up was in progress. The AIMS
demonstrated the best psychometric
properties in a systematic review
of motor assessments for preterm
infants.6 The BSID-II-NL is moderately reliable and valid according
to studies on the instruments psychometric qualities, but it has low
reliability until the age of 12
months.12,13 The test construction of
the AIMS is based on the dynamical
systems theory of motor development,11,14 whereas the PDI is based
on the traditional neuromaturational
concept.9
The neuromaturational theory proposes that changes in gross motor
skills during infancy result only from
the neurological maturation of the
central nervous system (CNS). The
dynamic motor theory considers
the CNS as one subsystem of many
that dynamically interact to develop
movements. Other elements that
explain movement changes are the
infants biomechanical and psychological factors and the nature of the
task or environment.11 Overall, the
AIMS is considered to have good
validity and high reliability and is
able to detect subtle changes in
movement quality.11,14,15 Responsiveness or sensitivity to change,
to our knowledge, has not been
documented. The data of both concurrent assessments enabled evaluation of intervention effects on
the intermediate time point of 12
months, in the light of the abovementioned need for sensitive assess-

November 2013

Evaluating Efficacy of Intervention on Motor Outcome

ment tools that measure change in
motor performance.
The purpose of this study was to
compare the AIMS and the PDI of
the BSID-II-NL in their ability to
evaluate the effects of an early neurobehavioral intervention program, the
IBAIP, on motor development in
infants with VLBW at 12 months CA.

Method
Participants and Procedure
The study sample consisted of
infants with VLBW of 12 months CA
participating in an RCT7 assessing
the effect of a neurobehavioral intervention program, the IBAIP.8 Two
level III hospitals with neonatal
intensive care unit facilities and all
5 city hospitals in Amsterdam, the
Netherlands, participated in this
RCT. After recruitment, 176 infants
with a gestational age (GA) of less
than 32 weeks or a birth weight less
than 1,500 g, or both, were included.
Exclusion criteria were severe congenital abnormalities of the infant,
severe physical or mental illness or
problems of the mother, non-native
families for whom an interpreter
could not be arranged, and participating in other trials on postdischarge management. After computergenerated randomization, stratified
for GA ( and 30 weeks) and
recruitment site with multiplets
assigned to the same group, 86 participants were assigned to the intervention group and 90 were assigned
to the control group. The study flow
diagram is presented in the Figure.
The intervention started a few days
before discharge. As at that point in
time neither the AIMS nor the PDI is
applicable, the standardized Infant
Behavioral Assessment16 (IBA) was
administered between 35 and 38
weeks postmenstrual age. The IBA
systematically observes and interprets 113 infant communicative
behaviors that are categorized
according to 4 subsystems: the autoNovember 2013

nomic system, the motor system, the

state system, and the attention/interaction system. Within each of the
4 subsystems, the behaviors are
interpreted as approach (stable/engagement), self-regulatory, or stress
(unstable/disengagement) behaviors.
In agreement with the higher biological risk of the IBAIP group, this
assessment showed that, at baseline,
infants in the intervention group
showed significantly less interaction
and more stress.7
The intervention period ended at
6 months CA. At 12 months CA,
6 months after the intervention had
ended, pediatric and developmental
assessments were performed during
1 visit at the follow-up clinic. First,
the standard clinical neurological
examination and the AIMS were
administered by 1 of 3 experienced
pediatricians. Subsequently, 1 of 2
developmental psychologists administered the BSID-II. The visit lasted
about 2 hours.
The AIMS was added to the assessment protocol 2 months after the
start of the follow-up at 12 months
CA. Therefore, not all infants
included in the original RCT were
assessed with the AIMS, and those
not assessed with the AIMS were
excluded from this study. Although
no interrater reliability was calculated, all assessors were trained
according to the standardized
instructions of both tests and were
blinded for group assignment.
The infants and the parents in the
intervention group received 1 intervention session shortly before discharge and 6 to 8 one-hour sessions
at home from an IBAIP-trained pediatric physical therapist. The control
group received standard care and
was referred to a non-IBAIP-trained
physical therapist if deemed necessary by the pediatrician. Written
informed consent was obtained from
the parents.
Volume 93

Measurements
Two standardized instruments documented the infants motor development. The AIMS11 is a measure of
infant gross motor development. It is
designed to measure motor skills
from term age to 18 months of age.
The test consists of 58 items divided
into 4 subscales: prone, supine, sit,
and stand. In each item, the qualitative aspects of the movement are
specifically described in terms of
weight-bearing surface of the body,
the posture necessary to achieve the
gross motor skill, and the antigravity
or involuntary movement performed
by the infant in the position. It has
been set as the norm on 2,202
infants born in the province of
Alberta, Canada. Raw total scores
and subscale scores can be converted to centile ranks and compared with the ranks of ageequivalent peers. Mildly delayed
motor development on the AIMS is
defined as a total score below the
10th percentile, and abnormal motor
development is defined as a score
below the 5th percentile. The AIMS
can be easily administered in clinical
settings; requires minimal handling,
and can be completed in 20 minutes.
The BSID-II-NL9 is used to assess
the mental and psychomotor development of children aged 1 to 42
months. It consists of mental, behavioral, and psychomotor scales.
Because of the aim of this article,
only the psychomotor scale is
described here. The 111 items of the
psychomotor scale measure fine and
gross motor skills. Depending on
the age and developmental level of
the infant, an age-appropriate set of
items is administered. Raw scores
can be converted in the PDI with, in
the population with normal motor
development, a mean (SD) of 100
(15). Mildly delayed motor development is defined as less than 85 points
(1 SD), and abnormal motor development is defined as less than 70
points (2 SD). The standardization
Number 11

Physical Therapy f

1477

Evaluating Efficacy of Intervention on Motor Outcome

315
eligible participants

176
randomized

86 intervention group infants

Follow-up at 6 mo (86)
PDI (BSID-II-NL) (86)

Follow-up at 12 mo (84)
withdrawn (1),
moved abroad (1)

139 excluded
refused to participate (38)
died (11)
child factors (12)
language reasons (11)
parental factors (12)
older brother/sister in trial (3)
participating in another trial (52)

90 control group infants

died before discharge (1)

Follow-up at 6 mo (85)
withdrawn (1), lost in follow-up (3)
PDI (BSID-II-NL) (83)

Follow-up at 12 mo (79)
withdrawn (2), lost in follow-up (1),
died (1), moved abroad (2)

PDI (BSID-II-NL) (83)

AIMS (58)

PDI (BSID-II-NL) (77)

AIMS (58)

Follow-up at 24 mo (83)

Follow-up at 24 mo (78)

moved abroad (1)

lost in follow-up (1)

PDI (BSID-II-NL) (75)

PDI (BSID-II-NL) (74)

Figure.
Study flow diagram. PDI (BSID-II-NL)Psychomotor Developmental Index (PDI) of the Bayley Scales of Infant DevelopmentDutch
Second Edition, AIMSAlberta Infant Motor Scale.

was based on the test results of 1,909

Dutch children and set into Dutch
norms in 2002. It takes about 45 minutes to administer the PDI.
Infant Behavioral Assessment
and Intervention Program
The IBAIP8 is a neurobehavioral
intervention program. To attend the
IBAIP training and become certified,
clinical experience with newborns
or young infants and knowledge of
1478

Physical Therapy

Volume 93

infant development and standardized testing are required. The IBAIP,

therefore, is accessible for pediatric
physical therapists. The IBAIP aims
to support the infants self-regulatory
competence and multiple developmental functions via responsive
parent-infant interactions. In practice, the interventionist teaches the
parent to recognize and interpret
the 3 different kinds of communicative behaviors of the infant in daily

Number 11

life activities as approach behaviors,

self-regulatory behaviors, or stress
behaviors to promote interactions
that match the infants needs.
Facilitation strategies are offered to
best support the infants neurodevelopmental progression and selfregulatory competence. As a result,
the infant is able to interact and
explore while maintaining stable
physiological and behavioral functioning. The facilitation strategies
November 2013

Evaluating Efficacy of Intervention on Motor Outcome

address environmental facilitation
(eg, visual and auditory input),
handling and positioning (eg, the
infants position in supine or prone),
and cue-matched facilitation (eg,
hand to mouth, foot bracing, or
hands to midline). Thus, the program supports the infants growth,
the infants motivation to explore,
and the possibility to learn from
information. For more detailed information about the content of the
IBAIP, the reader is referred to Hedlund8 and Hedlund and Tatarka.16
Data Analysis
Data were analyzed using the SPSS
computer program, version 16.0
(SPSS Inc, Chicago, Illinois). Differences in sociodemographic and perinatal characteristics between participants and nonparticipants and
motor outcomes of the AIMS and
PDI among participants and between
intervention and control groups were
analyzed using the independentsamples t test and the chi-square test,
when appropriate. The Pearson (r)
correlation coefficient was used to
assess the relationship between the
raw scores of the PDI and the AIMS
total and subscale scores. Multiple
linear regression analyses were used
to assess the effect of the intervention on the PDI and AIMS total and
subscale scores. We adjusted scores
for the following variables that differed at baseline: IBA approach and
stress behavior, sex, oxygen therapy
28 days, surfactant treatment, and
continuous positive airway pressure
(CPAP). The adjusted scores were
included in the linear regression
model as covariates because no collinearity among these factors was
found. An alpha level of .05 was considered for all tests of significance.
Effect sizes (ESs) were calculated
as the adjusted mean difference
between the intervention and control groups divided by the standard
deviation of the total group. To interpret the ES, we used Cohens criteria: 0.2small effect, 0.5modNovember 2013

erate effect, and 0.8large effect.17

To compare motor outcomes
between the intervention and control groups, binary logistic regression was performed, adjusted for
baseline differences with the AIMS
and PDI outcomes (normal versus
mildly abnormal motor development) as dependent variables, and
odds ratios (ORs) were calculated.
Role of the Funding Source
The study was supported by grants
from Innovatiefonds Zorgverzekeraars (project number 576) and
ZonMw (Zorg Onderzoek Nederland) (project number 62200032).
There were no conflicts of interest
with the study sponsors. The sponsors had no involvement in study
design; collection, analysis, or interpretation of the data; writing of the
report; or the decision to submit the
manuscript for publication.

Results
Sociodemographic and
Perinatal Factors
At 12 months CA, 116 of the 176
infants participating in the RCT were
assessed with the PDI and the AIMS.
They were equally divided between
the intervention group (n58) and
the control group (n58). Of the
60 infants who formed no part of
this study, 47 infants were assessed
only with the PDI and 13 infants
did not participate in the assessment
at 12 months CA. The participants
(n116) did not differ from the nonparticipants (n60) with respect to
sociodemographic and perinatal factors, except for 4 factors. Compared
with the nonparticipants, the participants had fewer approach behaviors
and more stress behaviors, more
low-educated fathers (46.5% versus
26.3%, P.011), less artificial ventilation (37.1% versus 53.3%, P.039),
a lower occurrence of ventricular
dilation (1.7% versus 8.3%, P.033),
and fewer septic periods (44.0% versus 60.0%, P.044).

Volume 93

A description of the sociodemographic and perinatal factors in the

intervention and control groups is
presented in Table 1. At baseline, the
infants in the intervention group
showed more stress behaviors and
fewer approach behaviors, as measured with the IBA. In accordance
with findings at 6 and 24 months7,10
and despite randomization, significantly more infants in the intervention group received respiratory
therapy, as indicated by surfactant
treatment, CPAP, and oxygen therapy 28 days, than in the control
group. In addition, there were significantly more boys in the intervention
group. Because of this imbalance
between the 2 groups, we adjusted
the outcomes for the IBA at baseline
and these perinatal factors.
Motor Outcomes and
Intervention Effects at
12 Months CA
The mean corrected test age at
motor assessments was 12 months
and 1 week. The total group of 116
infants with VLBW had a mean (SD)
AIMS total score of 48.1 (8.6) and a
mean (SD) PDI of 99.8 (15.0). Of the
116 infants with VLBW, abnormal
motor development was determined
in 20.7% of the infants based on the
AIMS versus 2.6% based on the PDI.
Mildly abnormal motor development
was determined in 31.9% of the
infants using the AIMS versus 8.6%
using the PDI. The PDI correlated
significantly with the AIMS total
score (r.726, P.001) and with the
AIMS subscales supine (r.412,
P.001), prone (r.626, P.001),
sit (r.582, P.001), and stand
(r.751, P.001).
Table 2 shows the motor outcomes
in the intervention and control
groups. The infants in the intervention group had significant higher
AIMS total and subscale sit scores
than the infants in the control group.
After adjustment for the baseline differences, a significant intervention
Number 11

Physical Therapy f

1479

Evaluating Efficacy of Intervention on Motor Outcome

Table 1.
Sociodemographic and Perinatal Characteristics of the Study Participantsa
Intervention Group
(n58)

Characteristic

Control Group
(n58)

Social background factors

Single-parent family, n (%)
Maternal age at date birth (y), X (SD)

9 (15.5)

5 (8.6)

.254

31.7 (5.0)

32.2 (5.0)

.642

Paternal age at date birth (y), X (SD)

36.1 (7.4)

35.7 (6.0)

.335

Mother born in the Netherlands, n (%)

33 (56.9)

38 (65.5)

.341

Father born in the Netherlands, n (%)

34 (58.6)

35 (60.3)

.850

Maternal education low, n (%)

26 (44.8)

23 (39.7)

.573

Paternal education low, n (%)

29 (50.0)

24 (41.4)

.265

Gestational age (wk), X (SD)

29.8 (2.1)

29.9 (2.0)

.440

Gestational age 28 wk, n (%)

12 (20.8)

6 (10.3)

.124

Perinatal factors

Birth weight (g), X (SD)

1,248 (338.6)

Small for gestational age,b n (%)

15 (12.9)

Sex: male, n (%)

37 (63.8)

Twins/triplets, n (%)

16 (27.6)/7 (12.1)

Antenatal steroid use, n (%)

Apgar score at 5 min, mean (SD)

1,315 (317.3)
8 (6.9)
26 (44.8)
15 (25.9)/1 (1.7)

.267
.103
.040*
.075

41 (70.7)

42 (72.4)

.833

8.5 (1.7)

8.6 (1.4)

.765

Surfactant, n (%)

21 (36.2)

11 (19.0)

.038*

Artificial ventilation, n (%)

24 (41.4)

19 (32.8)

.336

CPAP, n (%)

51 (87.9)

40 (69.0)

.013*

Oxygen support 28 d PMA, n (%)

22 (37.9)

10 (17.2)

.013*

Indomethacin use, n (%)

12 (20.7)

5 (8.6)

.066

Septic periods before discharge, n (%)

30 (51.7)

21 (36.2)

.092

Intraventricular hemorrhagec grade 1, 2/3, 4, n (%)

Periventricular leucomalacia,d n (%)

10 (17.2)/3 (5.2)

8 (13.8)/2 (3.4)

.859

8 (13.8)

6 (10.3)

.569

IBA total approach, X (SD)

3.1 (1.7)

3.9 (1.8)

.005*

IBA total regulation, X (SD)

12.4 (3.3)

12.7 (3.1)

.500

IBA total stress, X (SD)

12.4 (3.3)

11.4 (2.6)

.003*

At discharge

Length of hospitalization (d), X (SD)

57 (29.4)

Weight (g), X (SD)

2,420 (430.5)

Oxygen supply at discharge, n (%)

4 (6.9)

49 (22.1)
2,339 (410.9)
2 (3.4)

.100
.297
.402

Differences in mean scores and proportions between the groups are analyzed using t tests or chi-square tests. CPAPcontinuous positive airway pressure,
IBA Infant Behavioral Assessment, PMApostmenstrual age. *P.05.
b
Small for gestational age was defined as 1 SD below mean Dutch reference data.
c
Intraventricular hemorrhage defined according to Papile et al.24
d
Periventricular leucomalacia is defined according to de Vries et al.25

effect was found on the AIMS total

score and all AIMS subscale scores,
as well as on the PDI. The AIMS total
score and the subscale scores supine
and sit had the largest ESs, which
were moderate according to Cohens
criteria. The ES of the PDI was small.
1480

Physical Therapy

Volume 93

The rates of mildly abnormal and

abnormal motor outcome scores in
both tests are presented in Table 3.
The AIMS classified 13.8% of the
infants in the intervention group as
having abnormal motor development versus 27.6% of the infants

Number 11

in the control group (P.071).

Adjusted for the baseline difference,
the OR for abnormal motor development on the AIMS was statistically
significant (OR0.17, P.012, 95%
confidence interval [CI]0.04 0.67).
The PDI classified 1.8% of the infants
November 2013

Evaluating Efficacy of Intervention on Motor Outcome

Table 2.
Comparison of Outcomes on the Alberta Infant Motor Scale (AIMS) and the Psychomotor Developmental Index (PDI) of the
Bayley Scales of Infant DevelopmentDutch Second Edition (BSID-II-NL) Between the Intervention (Infant Behavioral Assessment
and Intervention Program) Group and Control Groupa

P1

Intervention
Group
(n58)
Adjusted
Mean (SE)

Control
Group
(n58)
Adjusted
Mean (SE)

Adjusted
Mean
Difference

P2

Effect
Size

0.316.57

.031*

50.6 (1.0)

44.4 (1.1)

5.7

.000*

0.72

1.0

0.422.31

.171

19.3 (0.5)

17.7 (0.5)

1.7

.024*

0.44

0.3

0.030.52

.079

9.1 (0.1)

8.6 (0.1)

0.5

.002*

0.59

10.6 (2.3)

0.8

0.081.52

.031*

11.5 (0.3)

10.3 (0.3)

1.2

.001*

0.64

9.1 (4.0)

0.8

0.722.06

.261

10.1 (0.5)

8.6 (0.5)

1.5

.039*

0.39

97.4 (14.8)

4.8

0.6510.40

.083

102.8 (2.0)

96.5 (2.0)

6.3

.032*

0.42

Intervention
Group
(n58)
X (SD)

Control
Group
(n58)
X (SD)

Mean
Difference

95% CI

Total score

49.8 (7.0)

46.4 (9.8)

3.4

Prone

19.1 (3.1)

18.1 (4.1)

Supine

9.0 (0.8)

8.7 (0.8)

11.4 (1.4)
9.9 (3.7)
102.2 (15.0)

Measure
AIMS

Sit
Stand
PDI (BSID-II-NL)
a

Differences in mean scores were analyzed using t tests. Multiple linear regression analyses were used to assess the effect of the intervention on the
developmental scores, adjusted for standardized Infant Behavioral Assessment approach and stress behavior at baseline, surfactant, oxygen therapy 28 d,
continuous positive airway pressure, and sex. P1P value uncorrected, P2P value corrected, SEstandard error, 95% CI95% confidence interval. *P.05.

in the intervention group as having

an abnormal motor development
versus 3.5% of the infants in the
control group (P.566). The OR for
abnormal motor development on the
PDI was not significant (OR0.21,
P.266, 95% CI0.013.30).

Discussion
This study demonstrates how 2
motor assessment tools, the AIMS
and the PDI, differ in evaluating
effects of an early intervention on
motor development in infants with
VLBW. On both tests, we found

intervention effects. However, the

AIMS detected more specific intervention effects than the PDI at
12 months CA. Significantly more
infants had an abnormal motor score
according to the AIMS norms than
according to the PDI norms. In addition, a significant reduction of abnormal motor outcome in the intervention group was found with the AIMS
but not with the PDI.
Using 2 instruments with different
test constructs enabled the evaluation of different aspects of motor

development. The PDI assesses gross

and fine motor development and has
the advantage of a large age range
(1 42 months) during which it can
be repetitively used. A limitation of
the PDI is an uneven distribution of
the different motor skill items. For
11 out of 15 items, the standing position is required at the age of 12
months, whereas only 2 items are
assessed in sitting position and
another 2 items are assessed in the
prone position.9 The AIMS has the
advantage of assessing motor development in 4 positions (supine,

Table 3.
Rates of Abnormal Outcomes on the Alberta Infant Motor Scale (AIMS) and the Psychomotor Developmental Index (PDI) of the
Bayley Scales of Infant DevelopmentDutch Second Edition (BSID-II-NL) in the Intervention Group (Infant Behavioral Assessment
and Intervention Program) and Control Groupa
Intervention Group
n (%)

Control Group
n (%)

Unadjusted OR
(95% CI)

P1

Adjusted OR
(95% CI)

P2

14 (24.1)

23 (39.7)

0.48 (0.221.08)

.075

0.24 (0.090.68)

.008*

8 (13.8)

16 (27.6)

0.42 (0.161.08)

.071

0.17 (0.040.67)

.012*

Mildly abnormal (85 points)

4 (6.9)

6 (10.3)

0.59 (0.181.92)

.381

0.46 (0.121.73)

.252

Abnormal (70 points)

1 (1.8)

2 (3.5)

0.49 (0.435.57)

.566

0.21 (0.013.30)

.266

Measure
AIMS
Mildly abnormal (P10)
Abnormal (P5)
PDI (BSID-II-NL)

Multiple logistic regression analyses were used to determine the intervention effect on the normal versus abnormal test outcomes, unadjusted and adjusted
for standardized Infant Behavioral Assessment approach and stress behavior at baseline, surfactant, oxygen therapy 28 d, continuous positive airway
pressure, and sex. P1P value unadjusted, P2P value adjusted, ORodds ratio, CIconfidence interval, P10below the 10th percentile, P5below the
5th percentile.

November 2013

Volume 93

Number 11

Physical Therapy f

1481

Evaluating Efficacy of Intervention on Motor Outcome

prone, sit, and stand), providing
gross motor subscales, and incorporates other developmental elements
such as the gravitational position of
the infant, weight bearing, and postural alignment, but lacks the fine
motor skills. Infants with VLBW are
known to experience difficulties in
these qualitative aspects of motor
development, such as reduced active
flexion power and discrepancies
between the active muscle power
and passive muscle tone.18,19 The latter influences the capability in independent sitting.3 The assessment in
4 positions may explain the reason
that the AIMS defines a higher proportion of infants as mildly abnormal
or abnormal in infants with VLBW
than the PDI.

be useful to elucidate specific effects

of the intervention on the motor
system. An ES on the subscales sit
and supine suggests that the neurobehavioral intervention enhanced
the infants postural control. The
IBAIP aims to support self-regulation.
The self-regulatory strategies involve
the motor system and focus on midline orientation, which may have
helped to gain more control over
posture and movements and thus
improved development in supine
and sitting postures. These selfregulatory strategies are of clinical
relevance because postural control,
a basic ingredient for motor development, often is described as one of
the specific motor problems in preterm infants.2,3

Our mean PDI scores are comparable with those of Westera et al,12
who also found a relative high score
(100.5) in a large group (n207) of
infants with VLBW at 12 months CA.
This finding supports the reports on
overestimation of the PDI in infants
with VLBW at 12 months CA.
Indeed, the rates of abnormal motor
development were lower using the
PDI than using the AIMS. Infants
with VLBW experience difficulties
with items that encompass trunk
control and trunk rotation and tend
to compensate with hyperextension.2,3 Therefore, they have fewer
problems with items in the standing
position than in the sitting position
or with making transits in or out of a
sitting or supine position. Using the
AIMS, there were significantly fewer
infants with abnormal motor development in the intervention group.
No such effect was found with the
PDI. This finding adds to our conclusion that the responsiveness of the
AIMS to detect an intervention effect
was better than that of the PDI.

We expected positive intervention

effects on motor outcomes at 12
months CA because previous significant intervention effects were found
with the PDI at 6 and 24 months
CA.7,10 At 12 months of age, the positive intervention effects were found
even in this subset (71.2%) of the
RCT. The reason for this reduction
was that follow-up was already in
progress when the AIMS was added
to the protocol.

In this study, positive effects were

found on all AIMS subscales, especially on the subscales sit and supine.
Effect sizes on these subscales can
1482

Physical Therapy

Volume 93

The AIMS total score correlated less

with the PDI score in our study
(r.73, P.001) compared to other
publications (r.90 and r.89,
P.01).13,14 The discrepancies may
have been due to the use of the
Dutch version of the BSID-II or to
methodological differences such as
the number of assessed infants at
12 months CA. In our study, we
assessed 116 infants, and other studies evaluated 45 and 48 infants.13,14
To our knowledge, no correlation
values between the subscales of the
AIMS and the PDI have been
described to date. Not surprisingly,
the strongest correlation was between
the PDI and the AIMS subscale stand
because, as mentioned above, for 11
out of 15 items on the PDI at the age

Number 11

of 12 months, the standing position

is required.
A significant limitation of this study
is that no interrater reliability was
calculated for the raters who scored
the AIMS and the BSID-II-NL. We can
report that the 3 assessors of the
AIMS and 2 assessors of the BSIDII-NL were trained according to the
standardized instructions of both
tests. However, reliability between
raters remains a limitation, and our
findings should be cautiously
generalized.
The relevance of adding the AIMS to
the Bayley Scales of Infant Development (BSID) assessments is supported by growing concerns about
the high scores of the third edition
of the BSID20 in high-risk populations.2123 This new edition of the
BSID will be available for Dutch children in 2014. The BSID-III separates
the psychomotor scale into a fine
motor scale and a gross motor scale,
but in comparison with the second
edition, the gross motor assessment
items are assessed only in the standing position at the age of 12 months.
Because the abilities in which infants
with VLBW experience difficulties,
such as trunk control and trunk rotation, are underrepresented in both
the BSID-II and the BSID-III, the current results using the BSID-II-AIMS
comparison also are applicable for
the BSID-III. Further research is warranted to explore the correlation
between the outcomes of the AIMS
and motor scale of the BSID-III, as
well as between AIMS scores during
infancy and motor outcomes at
school age.

Conclusion
This study demonstrated that both
the AIMS and the PDI of the BSID-II
are able to evaluate the effect of
early neurobehavioral intervention
on motor development in infants
with VLBW at 12 months CA. However, the responsiveness of the AIMS
November 2013

Evaluating Efficacy of Intervention on Motor Outcome

to detect an effect of this intervention was better than that of the PDI.
Effects on the AIMS subscales sit and
supine pointed toward improved
postural control due to the intervention. In line with recent publications, we recommend caution with
monitoring infants with VLBW only
with the PDI and advise using both
the AIMS and the PDI to evaluate
intervention effects on motor development at 12 months CA.
Mrs Van Hus, Dr Jeukens-Visser, Dr Koldewijn, Dr Kok, Dr Nollet, and Dr Van
Wassenaer-Leemhuis provided concept/
idea/research design. Mrs Van Hus, Dr Koldewijn, Dr Van Sonderen, and Dr Van
Wassenaer-Leemhuis provided data collection. Mrs Van Hus and Dr Jeukens-Visser provided data analysis. Mrs Van Hus, Dr JeukensVisser, and Dr Van Wassenaer-Leemhuis
provided writing. Dr Kok, Dr Nollet, and Dr
Van Wassenaer-Leemhuis provided project
management. Dr Koldewijn, Dr Kok, Dr Nollet, and Dr Van Wassenaer-Leemhuis provided fund procurement and facilities/equipment. All authors provided consultation
(including review of manuscript before
submission).
The study was approved by the medical ethics committees of the 2 level III hospitals and
all 5 city hospitals in Amsterdam, the Netherlands. The Medical Ethics Committee of
Academic Medical Centre, Amsterdam, the
Netherlands, approved the follow-up study.
The study was supported by grants from
Innovatiefonds Zorgverzekeraars (project
number 576) and ZonMw (Zorg Onderzoek
Nederland) (project number 62200032).
The trial is registered with controlledtrials.com (ISRCTN65503576).
DOI: 10.2522/ptj.20120460

November 2013

References
1 De Kieviet JF, Piek JP, Aarnoudse-Moens
CS, et al. Motor development in very preterm and very low-birth-weight children
from birth to adolescence. JAMA. 2009;
302:22352242.
2 Pin TW, Darrer T, Eldrigde B, et al. Motor
development from 4 to 8 months corrected age in infants born at or less than 29
weeks gestation. Dev Med Child Neurol.
2009;51:739 745.
3 De Groot L, Hopkins B, Touwen BC. Muscle power, sitting unsupported and trunk
rotation in preterm infants. Early Hum
Dev. 1995;43:37 46.
4 Wang JC, McGlynn A, Brook RH, et al.
Quality of care indicators for the neurodevelopmental follow-up of very low birth
weight children: results of an expert process. J Pediatr. 2006;117:2080 2092.
5 Spittle AJ, Orton J, Doyle LW, et al. Early
developmental intervention programs
post hospital discharge to prevent motor
and cognitive impairments in preterm
infants. Cochrane Database Syst Rev.
2007;18:CD005495.
6 Spittle AJ, Doyle LW, Boyd RN. A systematic review of the clinimetric properties
of neuromotor assessments for preterm
infants during the first year of life. Dev
Med Child Neurol. 2008;50:254 266.
7 Koldewijn K, Wolf MJ, Van Wassenaer AG,
et al. The Infant Behavioral Assessment
and Intervention Program for very low
birth weight infants at 6 months corrected
age. J Pediatr. 2009;154:3338.
8 Hedlund R. The Infant Behavioral Assessment and Intervention Program. 1998.
Available
at:
http://www.ibaip.org.
Accessed February 15, 2012.
9 Van der Meulen BF, Ruiter SAJ, Lutje Spelberg HC, et al. BSID-II-NL: Dutch Manual.
Lisse, the Netherlands: Swets Test Publishers; 2002.
10 Koldewijn K, Van Wassenaer AG, Wolf MJ,
et al. A neurobehavioral intervention and
assessment program in very low birth
weight infants: outcome at 24 months.
J Pediatr. 2010;156:359 365.
11 Piper MC, Darrah J. Motor Assessment of
the Developing Infant. Philadelphia, PA:
WB Saunders Co; 1994.
12 Westera JJ, Houtzager BA, Overdiek B,
et al. Applying Dutch and US versions of
the BSID-II in Dutch children born preterm leads to different outcomes. Dev Med
Child Neurol. 2008;50:445 449.

Volume 93

13 Ruiter SAJ, Lutje Spelberg HC, van der

Meulen BF, et al. The BSID-II-NL: construction, standardization, and instrumental
utility. D J Psychol. 2008;64:15 40.
14 Jeng SF, Yau KI, Cheb LC, Hsiao SF.
Alberta Infant Motor Scale: reliability and
validity when used on preterm infants in
Taiwan. Phys Ther. 2000;80:168 178.
15 Almeida KM, Dutra MV, Mello RR, et al.
Concurrent validity and reliability of the
Alberta Infant Motor Scale in premature
infants. J Pediatr (Rio J). 2008;84:442 448.
16 Hedlund R, Tatarka M. The Infant Behavioral Assessment. 1988. Available at:
http://www.ibaip.org. Accessed February
15, 2012.
17 Cohen J. Statistical Power Analysis for
the Behavioral Sciences. Hillsdale, NJ:
Lawrence Erlbaum Associates; 1988.
18 Dewey D, Creighton DE, Heath JA, et al.
Assessment of developmental coordination disorder in children born with
extremely low birth weights. Dev Neuropsy. 2011;36:4256.
19 De Groot L, Van De Hoek AM, Hopkins B,
Touwen BC. Development of relationship
between active and passive muscle power
in preterms after term age. Neuropediatrics. 1992;23:298 305.
20 Bayley N. Bayley Scales of Infant and
Toddler Development. 3rd ed. San Antonio, TX: Harcourt Assessment; 2006.
21 Moore T, Johnson S, Haider S, et al. Relationship between test scores using the second and third editions of the Bayley scales
in extremely preterm children. J Pediatr.
2012;100:553558.
22 Anderson PJ, De Luca CR, Hutchinson E,
et al. Underestimation of developmental
delay by the new Bayley-III scale. Arch
Pediatr Adolesc Med. 2010;164:352356.
23 Vohr BR, Stephens BE, Higgins RD, et al.
Are outcomes of extremely preterm
infants improving? Impact of Bayley
assessment on outcomes. J Pediatr. 2012;
161:222228.
24 Papile LA, Munsick-Bruno G, Schaefer A.
Relationship of cerebral intraventricular
hemorrhage and early childhood neurologic handicaps. J Pediatr. 1983;103:272
277.
25 de Vries LS, Eken P, Dubowitz LM. The
spectrum of leukomalacia using cranial
ultrasound. Behav Brain Res. 1992;49:
1 6.

Number 11

Physical Therapy f

1483

Copyright of Physical Therapy is the property of American Physical Therapy Association and
its content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.