Acta Pñ diatr 89: 1456± 61. 2000
Early motor development of premature infants with birthweight less
than 2000 grams
SJ Pedersen, K Sommerfelt and T Markestad
Department of Pediatrics, University of Bergen, Norway
Pedersen SJ, Sommerfelt K, Markestad T. Early motor development of premature infants with
birthweight less than 2000 grams. Acta Pædiatr 2000; 89: 1456–61. Stockholm. ISSN 0803–5253
The aim was to assess motor function during infancy in order to predict later function, mainly
cerebral palsy. The neuromotor development of a population-based cohort of 209 of 236 (89%)
survivors with a birthweight less than 2000 g was assessed using the Infant Neurological
International Battery (INFANIB) and detailed neurological assessment. The infants were classiŽ ed
as being normal, dystonic, hypotonic or having suspected cerebral palsy (CP) at 4, 7, 13 and 18
mo corrected age if birthweight was less than 1500 g (n = 119) and at 7 and 13 mo if birthweight
was 1500–1999 g (n = 90). Those with dystonia or suspected CP were followed until diagnosed as
normal or as having CP after at least 36 mo of age. Fourteen (7%) Ž nally developed CP. Motor
function at 4 mo was inaccurate in predicting function at 7 mo and later. All who were normal at
7 mo remained normal in the follow-up period. Eight of 65 who were dystonic at 7 mo developed
suspected CP, and three judged as suspected CP were eventually normal. The 14 who developed
CP were judged as suspected CP (n = 5) or dystonic (n = 8) and one as hypotonic at 7 mo of age.
Conclusions: The speciŽ city of motor evaluation at 7 mo corrected age regarding CP is unsatisfactory,
since dystonia at this age is most often transient. A normal neuromotor assessment at 7 mo is highly
predictive of subsequent normal motor function.
Key words: Cerebral palsy, dystonia, low birthweight, outcome, prematurity
Svein Junker Pedersen, Department of Pediatrics, Barneklinikken, N-5021 Haukeland Sykehus,
Norway (Tel. + 47 55 975180, fax. + 47 55 975147)
Ó
For children with increased risk of neurodevelopmental
deŽ ciencies, such as preterm and low birthweight
infants, it is desirable to make early predictions with
regard to outcome. This is important for the family, for
the researchers designing appropriate follow-up and
intervention programmes, and for the paediatric neurologist who has to give meaningful feedback to obstetricians and neonatologists. Several studies have tried to
identify infants at particularly high risk of neurological
damage (1, 2). Few studies however, have been population-based, and many of them were published before the
introduction of modern neonatal intensive care, which
resulted in much higher survival rates in preterm babies
(3, 4).
In 1972, Drillien described transitory dystonia of the
low birthweight premature infant as a common developmental deviation of motor function during the Ž rst
year of life. The typical motor features described were
increased extensor tone of the trunk and lower
extremities, increased adductor tone in the lower
extremities, persistent primitive re exes, head lag on
traction and delayed supportive responses (3). The
symptoms may vary, and usually disappear gradually
between 8 and 12 mo of age. In some children the
symptoms will not disappear, and they will eventually
be diagnosed as having spastic cerebral palsy (3–8).
2000 Taylor & Francis. ISSN 0803-525 3
The purpose of the present study was to describe the
pattern of motor development in preterm infants with
birthweights less than 2000 g prospectively in a
population-based cohort, and to assess the predictive
value of a normal or a deviant pattern of motor function
at two speciŽ c ages during the Ž rst year of life.
Material and methods
All liveborn infants with birthweights less than 1500 g
born during the period 1.1.86 to 31.12.89 and all with a
birthweight of 1500–1999 g born during half that period
(1.1.88 to 31.12.89) having parents living in Hordaland
county, Norway, were eligible for the study. Hordaland
county has a population of approximately 416,000,
which is 10% of the population of Norway, and the
demographic characteristics of the county are similar to
those of Norway as a whole. The annual number of
births during these two periods was 5,643–5,083. All
neonates with birthweights less than 2000 g were
admitted to the only neonatal intensive care unit in the
county.
During 1986–89 the perinatal treatment protocol did
not change. Indications and methods for ventilatory
therapy and withholding treatment remained un-
Motor developmen t of premature infants
ACTA PÆDIATR 89 (2000)
1457
Table 1. Survival and loss to follow-up among the liveborn infants with birthweight s under 2000 g. Number of cases (per cent).
Birthweight (g)
Birthweight group
a
Total liveborn
Boys/girls
Neonatal death
Neonatal survivors
Dead > 4 wk < 6 moc
Multiple malformation/chromos.aberrationd
Eligible survivors
Lost to follow-up because they moved
Lost to follow-up because they refused
Lost to follow-up, followed elsewhere
Lost to follow-up, other causes
Assessed of those eligible e
Boys/girls
Birthweight 1 SD
Gestational age 1 SD
a
b
c
d
e
Total
275
142/133
24
251 (91)
7
8
236
7
5
9
6
209 (89)
110/99
1423 356
31 3
< 1000
49
21/28
17 b
32 (65)
1
1
30
3
0
1
0
26 (87)
11/15
819 119
26 2
1000–1499
111
56/55
6
105 (95)
3
2
100
3
1
2
1
93 (93)
46/47
1263 142
30 2
1500–1999
115
65/50
1
114 (99)
3
5
106
1
4
6
5
90 (85)
53/37
1764 145
33 2
1.1.86–31.12.87, birthweight < 1500 g, 1.1.88–31.12.89, birthweight < 2000 g.
Two children had lethal malformations, the others died of causes related to prematurity and asphyxia.
One child died of causes related to prematurity; Ž ve died of SIDS between 3 and 6 mo of age, one died of BPD/cerebral ishaemia.
Six children died between 5 wk and 2 y of age. One child developed cerebral palsy after physical abuse; one child with Vater syndrome survived.
Percent calculated from the number of eligible survivors.
changed. Prenatal steroids were given whenever possible if gestational age was less than 34 wk, and postnatal
steroids for bronchopulmonary dysplasia were given at
approximately 2 wk of age if the baby was still on a
ventilator. Surfactant treatment was not introduced.
In the follow-up programme, the infants with birthweights less than 1500 g were seen at 4, 7, 13 and 18 mo
of corrected age, i.e. age after expected term date.
Children with a birthweight of 1500–1999 g were seen
at 7 and 13 mo corrected age. If these last infants had
not reached normal developmental milestones according to the Denver developmental screening test (9),
including independent walking, and had normal neurological examinations at 13 mo, they were seen at 18 mo.
Standard neurological examination was used instead of
INFANIB after 18 mo of age. The Ž rst author followed
children classiŽ ed as dystonic at 18 mo of age until
maximum 36 mo of age (deŽ ned as Ž nal outcome).
Information regarding further development of those
with hypotonia was partly gathered through a later
follow-up study. In this study, a subgroup of children
from the present study (all with birthweight less than
1500 g, born 1986–1988) were examined at 5 y of age
(10). For the rest, information was gathered from a
postal questionnaire or telephone interview when the
children were 9–11 y of age. Postal questionnaires,
telephone interviews or health record information were
also gathered for those who were eligible for the study
but lost to follow-up. Children with suspected cerebral
palsy (CP) were also referred to the habilitation centre,
and those with established cerebral palsy were seen
there regularly during childhood.
At each follow-up visit the infants were subjected to a
standardized and detailed clinical examination by S.J.P.
or T.M., both senior paediatricians experienced in
neurological and developmental assessments. Neurological status was assessed using the INFANIB (Infant
Neurological International Battery) (11–13), and a
standard neurological examination with emphasis on
primitive re exes, deep tendon re exes and quality of
movements. INFANIB was not scored numerically, but
on the basis of these examinations the infant was on
each visit classiŽ ed into one of four possible groups
with regard to motor function (motor classiŽ cation);
normal, dystonic, hypotonic or suspected cerebral palsy
(CP). An infant with no obvious deviation in motor
function was deŽ ned as normal. Dystonia was deŽ ned
as increased extensor tone of the lower extremities and/
or trunk, and increased adductor tone in the lower
extremities. In the INFANIB these deviations were
detected through the items “heel to ear”, “popliteal
angle”, “leg abduction”, “ankle dorsi extion” or
“standing”. Increased deep tendon re exes were often
seen in these infants, but it was not a requirement for
giving the diagnosis. Children with marked dystonia
also had increased extensor tone in the trunk and neck
disclosed in items like “tonic labyrinth supine”, “pull to
sitting”, “sitting”, “prone position” and “scarf sign”, or
had immature movements during “body derotative” or
“body rotative”. Hypotonia was deŽ ned as decreased
muscle tone, increased mobility of joints, delayed
support and balance reactions, and delayed motor
milestones expressed in items like “heel to ear”,
“popliteal angle”, “leg abduction”, “scarf sign”, “ankle
dorsi ection”, “pull to sitting”, “sitting”, “weight
bearing”, “body rotative” and “forward, sideways and
backwards parachute” in INFANIB.
CP was deŽ ned as “a disorder of movement and
posture due to a defect or lesion of the immature brain”
and characterized by a neurological deŽ cit with
1458
SJ Pedersen et al.
ACTA PÆDIATR 89 (2000)
Fig. 1. Motor developmen t of premature infants with birthweight s less than 1500 g and 1500–1999 g, respectivel y (—^—normal; --&-dystonia; ——~——hypotonia ; —&—suspected CP).
increased tone, abnormal re ex pattern and persistently
abnormal paretic patterns of posture and movement
(14). The various cerebral palsy syndromes were
classiŽ ed as described by Hagberg et al. (15). The term
“suspected CP” was used in the presence of the same
symptoms until the age of at least 36 mo if there was any
doubt as to Ž nal outcome.
Each infant was classiŽ ed according to the most
deviant motor function observed during one or more
follow-up visits during the Ž rst 18 mo of corrected age
(Table 2). Most deviant was suspected CP followed by
dystonia and hypotonia.
At 7 mo corrected age the infants were tested with
“the Fagan tests of Infant Intelligence” (16), which is a
paired comparison test of visual novelty preference
pertaining to the infants’ ability to process information.
At 13 mo corrected age the infants were assessed with
Table 2. ClassiŽ cation of motor function. The children were grouped
according to the most deviant function on at least one visit during the
Ž rst 18 mo. Number of cases (per cent).
Birthweight (g)
Motor classiŽ cation
Total no.
209
< 1000
26
Normal
Dystonia
Hypotonia
Suspected CP
111 (53)
61 (29)
21 (10)
16 (8)
11
9
2
4
(42)
(35)
(8)
(15)
1000–1499 1500–1999
93
90
43
33
10
7
(46)
(35)
(11)
(8)
57
19
9
5
(63)
(21)
(10)
(6)
the Bayley scales of infant development (17). A
psychometrist who was unaware of the background of
the baby or the assessment by the physician performed
both the Fagan and Bayley tests. At each visit the
infants were also seen by a physiotherapist, who
examined the baby according to the Bobath method
and gave advice and support to the family. Prophylactic
or therapeutic interventions were not initiated, however,
unless suspected cerebral palsy was diagnosed.
The INFANIB has not been validated for Norwegian
infants. As part of the present study, 26 infants (14 boys,
12 girls) born at term with a mean birthweight of
3,657 g (SD 468 g) and mean gestational age of 39 wk
(SD 1 wk) were examined at 7 mo of age. Two of the
infants (2%) were judged to be mildly dystonic, while
the others were normal according to the American
standard. All 26 children had a subsequent normal
development at 13 mo and 5 y of age.
The SPSS for Windows was used for statistical
analysis, and differences were tested for statistical
signiŽ cance using the chi squared test and ANOVA
analysis of variance. Values of p < 0.05 were considered statistically signiŽ cant.
Results
Of 275 liveborn preterm infants born in the deŽ ned
period, 236 fulŽ lled the criteria for follow-up (Table 1).
1459
Motor developmen t of premature infants
ACTA PÆDIATR 89 (2000)
Table 3. Outcome at 7 mo corrected age and Ž nal outcome compared with motor classiŽ cation at corrected age of 4 mo for infants with
birthweight s less than 1500 g (110 of 119a infants seen at both 4 and 7 mo corrected age). Number of cases.
ClassiŽ cation at 7 mo
Motor classiŽ cation at 4 mo
Normal
Dystonia
Hypotonia
Suspected CP
a
b
c
d
Final outcome ( 36 mo)
Total
Normal
Dystonia
Hypotonia
Suspected CP
Normal
CP
79
25
4
2
53
3
1
0
20b
18c
1
0
6
0
2
0
0
4d
0
2
76
20
4
0
3
5
0
2
Three children were not seen at 4 mo of age, six children were not seen at 7 mo of age. All these nine were judged as normal.
Three children were later found to have suspected CP (diplegia).
Two children were later found to have suspected CP (diplegia).
One child with suspected CP (diplegia) was judged as normal by 2 y of age.
None of those who died after discharge from the
hospital showed signs of neurological damage at
discharge. Of the 236 eligible infants, 27 were not seen
for various reasons (Table 1). For 10 of these the last
information was from hospital charts and all were
judged to be normal on the last visit at 4–30 mo
corrected age, except one child who had “Apple-peel”
syndrome. The remaining 17 children were normal at 9–
11 y of age judged from postal questionnaires or
telephone interviews with the parents.
Fifty-three percent of the infants followed according
to the programme were found to have a normal motor
function at every visit (Table 2, Fig. 1). There was a
tendency towards an increased proportion of deviant
motor function with decreasing birthweight, but the
differences were not statistically signiŽ cant (Table 2). A
deviant motor function was most commonly seen at 7
mo, and dystonia was the predominant dysfunction (Fig.
1, Table 2).
The subsequent return to a normal motor function at
18 (birthweight less than 1500 g) and 13 mo of age
(birthweight 1500–1999 g), respectively, was due to a
reduction in the number of infants with dystonia (Fig.
1). At 7 mo of age, 41 of 113 (36%) infants with
birthweights less than 1500 g were dystonic compared
with 9 of 112 (8%) at 18 mo corrected age. For those
with birthweights of 1500–1999 g, 24 of 85 (28%) were
dystonic at 7 mo of age compared with 4 of 90 (4%) at
13 mo of age. Of 16 infants classiŽ ed as hypotonic at 7
mo of age, 3 were normal at 13 mo of age, 2 were lost to
follow-up after 13 mo of age and 11 were still hypotonic
at both 13 and 18 mo of age.
Fourteen children were given a Ž nal diagnosis of CP,
eight (57%) had spastic diplegia, three (21%) had
hemiplegia and three (21%) tetraplegia. Nine (64%) had
birthweights less than 1500 g and Ž ve (36%) had
birthweights 1500–1999 g. In the period 1.1.88 to
31.12.89, when the study included all with birthweight
less than 2000 g, 11 (7%) of 153 children developed CP,
i.e. six (4%) of those with birthweight less than 1500 g
and Ž ve (3%) with birthweight 1500–1999 g.
Two children, both with spastic tetraplegia, were
diagnosed as suspected CP before discharge from
hospital, 3 at 7 mo, 5 between 7 and 18 mo of age
and 4 between 18 and 36 mo of age. All were eventually
given the diagnosis of CP. In addition, three children
were assessed as having suspected spastic diplegia
between 7 and 18 mo of age, but were judged normal at
24 mo of age. For the nine children with suspected CP
who were Ž rst diagnosed as suspected CP later than 7
mo, eight were Ž rst classiŽ ed as dystonic and one as
hypotonic. Only one of those with a Ž nal diagnosis of
spastic diplegia or hemiplegia had a frank mental
retardation. This was evaluated at the habilitation centre
when the children were between 4 to 6 y of age. At the
same age, 10 children with spastic diplegia or hemi-
Table 4. Outcome at 13 mo corrected age and Ž nal outcome compared with motor classiŽ cation at a corrected age of 7 mo for infants with
birthweight s less than 2000 g. (Total 198 of 209a infants seen, 113 of 119 infants with birthweight s less than 1500 g and 85 of 90 infants with
birthweight s of 1500–1999 g). Number of cases.
ClassiŽ cation at 13 mo
Motor classiŽ cation at 7 mo
Normal
Dystonia
Hypotonia
Suspected CP
a
b
c
d
e
Final outcome ( 36 mo)
Total
Normal
Dystonia
Hypotonia
Suspected CP
Normal
CP
111
65
16
6
106
34
3
0
0
20b
0
0
5
4
13d
0
0
7c
0
6e
111
57
15
1
0
8
1
5
Eleven children were not seen at 7 mo of age, but all were normal at 13 mo of age.
Three children were later found to have suspected CP before 24 mo of age.
Two children with suspected CP (diplegia) were judged as normal by 24 mo of age.
One child with truncal hypotoni a was judged as having suspected CP (diplegia) at 28 mo of age.
One child with suspected CP (diplegia) was judged as normal by 24 mo of age.
1460
SJ Pedersen et al.
ACTA PÆDIATR 89 (2000)
Table 5. Bayley mental (MDI) and Bayley motor (PDI) scores (mean 1SD) at 13 mo and Fagan score at 7 mo corrected age in relation to
motor classiŽ cation. Infants with birthweight s less than 1500 g (total 87 of 119) and birthweight s 1500–1999 g (total 82 of 90). Scores for each
deviant group are compared statisticall y with the normal group. T-tests were used.
Motor classiŽ cation
Birthweight < 1500 g
Normal
Dystonia
Hypotonia
Suspected CPc
Birthweight 1500–1999 g
Normal
Dystonia
Hypotonia
Suspected CPd
a
b
c
d
No. examined/
total
Bayley/MDI a
(mean)
Bayley/PDI b
(mean)
p-value
p-value
No. examined/
total (mean)
Bayley/PDI b
(mean)
p-value
41/54
29/42
11/12
6/11
108
98
95
96
14
16
17
14
ns
ns
ns
100
90
81
79
16
15
12
1
0.048
0.005
0.02
26/54
24/42
9/12
2/11
56.4
58.0
55.6
62.4
8.4
5.5
5.2
2.1
ns
ns
ns
50/57
19/19
9/9
4/5
108
106
97
101
12
9
16
12
ns
ns
ns
102
102
86
73
14
13
19
14
ns
0.031
0.003
31/57
12/19
5/9
3/5
57.0
55.8
60.4
58.3
7.1
4.0
4.6
4.3
ns
ns
ns
MDI: Mental developmen t index.
PDI: Psychomotor developmen t index.
Two children with severe suspected CP(tetraplegia) were not examined.
One child with severe suspected CP(tetraplegia) was not examined.
plegia were walking independently, one of them with
aides. The children with spastic tetraplegia were
wheelchair-dependent.
Neurological assessment at 4 mo corrected age was
inaccurate in predicting normal or deviant motor
function at 7 mo, and in predicting Ž nal outcome (Table
3). However, none of those who were considered
normal at 7 mo subsequently developed dystonia or
cerebral palsy (Table 4).
Eight of the 65 (12%) who were dystonic and 1 of the
16 (6%) who were hypotonic at 7 mo developed
suspected CP, while the others eventually had a normal
motor development. Of the 14 children who eventually
developed CP, 5 (36%) were identiŽ ed as suspected CP
at or before 7 mo corrected age.
There were no signiŽ cant differences in the mean
scores on the Fagan test between the different motor
classiŽ cation groups (Table 5). At 13 mo corrected age,
the mean Bayley PDI was signiŽ cantly lower for those
who had been diagnosed as dystonic, hypotonic or with
suspected cerebral palsy, compared with those who
persistently had been judged as normal (Table 5).
Discussion
The present study conŽ rms that low birthweight infants
commonly show a pattern of motor development during
their Ž rst year of life which is different from that
expected for infants born at term. Furthermore, few of
these infants with deviant motor development eventually develop cerebral palsy. In the present study, 47%
of those with birthweights less than 2000 g were
classiŽ ed as having a deviant motor function at least
on one occasion, and most commonly at the age of 7 mo
after expected term. A higher proportion of those with
the lowest birthweights tended to be classiŽ ed as having
transient dystonia. The population-base d and prospec-
tive design, relatively high survival rate and high
follow-up rate suggest that the Ž ndings in the present
study are representative for survivors of prematurity in
countries offering advanced neonatal care, but before
the introduction of surfactant treatment.
Several other authors have found that transient
dystonia was most commonly detected at 3–5 mo of
age, with a subsequent decrease in motor abnormalities
from 8 mo until normal Ž ndings at 12–18 mo of age
(3, 4, 6, 8, 18). In the present study, infants with birthweights of 1500–1999 g were not examined at 4 mo
corrected age, but for those with lower birthweights
dystonia was most common at 7 mo corrected age and
subsequently decreased markedly over the next 6 mo.
Still, a few infants who later had a normal outcome were
dystonic as late as 18 mo after expected term.
In the present study, the signiŽ cantly lower mean
Bayley psychomotor score at 13 mo of age for those
with dystonia compared to the normal group without
motor problems indicates concern regarding the signiŽ cance of transient dystonia for future development.
However, a follow-up study at 5 y of age of a subgroup
of the present cohort showed that transient dystonia was
not associated with cognitive or behavioural deŽ cits,
but there was a tendency, although not statistically
signiŽ cant, for a higher rate of minor motor coordination problems in this group (10).
This Ž nding is in accordance with the few other
relatively recent studies (19, 20) and in contrast to
studies on infants born in the late 1960s and early 1970s
in which transient dystonia was associated with an
increased frequency of neurodevelopmental abnormalities (21, 22).
Several authors have tried to predict long-term
outcome in preterm infants based on evaluations during
infancy. The designs of these studies have differed in
respect to the methods used and to the age of the infants
when making the predictions and deŽ ning the outcome
ACTA PÆDIATR 89 (2000)
(1, 2). Most of the studies report a high sensitivity but a
very low speciŽ city in predicting later abnormal outcome when the predictive assessments are made in the
neonatal period, or at 4–8 mo of age. Both Harris et al.
(1) and Swanson et al. (2) described low speciŽ city of
transient dystonia, which is in accordance with the
present study. Amiel-Tison claimed that it was not
possible to make a conŽ dent differentiation between
transient dystonia and cerebral palsy at 5–8 mo of age
(6), a Ž nding that was also conŽ rmed in the present
study.
In most follow-up studies the children included have
a birthweight less than 1500 g. In this weight group the
typical CP frequency is approximately 7% (23), as
found in the present study. In our study an interesting
Ž nding was that nearly half the children with suspected
CP in the cohort born in 1988–1989 had a birthweight
between 1500 and 1999 g. This is an important Ž nding
and may be one reason for including this weight group
too in follow-up programmes.
Only 36% of the children who developed suspected
CP were identiŽ ed before or at 7 mo of age in the
present study. A more recently introduced method
called general movements (GM) may enable detection
of abnormal motor outcome at an earlier age than
INFANIB, but the method is time-consuming and needs
further validation (24).
The essential Ž ndings in the present study were that
the future outcome is difŽ cult to assess at 4 mo
corrected age in preterm infants if severe cerebral
damage is not evident in the neonatal period. Relatively
few of those with deviant motor function at 7 mo will
develop cerebral palsy, but those who do cannot be
reliably identiŽ ed at this age. With a normal motor
function at 7 mo after expected term cerebral palsy is
very unlikely to develop, allowing reassurance to the
family for their child and a less extensive subsequent
follow-up programme for busy healthcare providers.
References
1. Harris SR, Swanson MW, Andrews MS, Sells CJ, Robinson NM,
Bennett FC, Chandler LS. Predictive validity of the “movement
assessment of infants”. J Dev Behav Pediatr 1984; 5: 336–42
2. Swanson MW, Bennett FC, Shy KK, WhitŽ eld M. IdentiŽ cation
of neurodevelopmenta l abnormality at four and eight months by
the movement assessment of infants. Dev Med Child Neurol
1992; 34: 321–37
3. Drillien CM. Abnormal neurologi c signs in the Ž rst year of life in
low-birthweight infants: possible prognosti c signiŽ cance. Dev
Med Child Neurol 1972; 14: 575–84
4. Tudehope DI, Burns YR, O’Callaghan M, Mohay H. Minor
neurologica l abnormalitie s during the Ž rst year of life in infants
of birth weight < 1500 g. Aust Paediatr J 1981; 17: 265–8
Motor developmen t of premature infants
1461
5. Brown JK, Drillien CM. The dystonic syndrome of the low
birthweigt infant. Arch Dis Child 1971; 46: 739–40
6. Amiel-Tison C. A method for neurologi c evaluatio n within the
Ž rst year of life. Cur Probl Pediatr 1976; 7: 1–70
7. Nelson KB, Ellenberg JH. Children who “outgrew” cerebral
palsy. Pediatrics 1982; 69: 529–36
8. Coolman RB, Bennett FC, Sells CJ, Swanson MW, Andrews MS,
Robinson NM. Neuromotor developmen t of graduates of the
neonatal intensive care unit: patterns encountered in the Ž rst two
years of life. J Dev Behav Pediatr 1985; 6: 327–30
9. Frankenburg WK, Dobbs JB, Fandal A. The revised Denver
Developmenta l Screening test manual. Denver: University of
Colorado Press; 1970
10. Sommerfelt K, Pedersen S, Ellertsen B, Markestad T. Transient
dystonia in non-handicappe d low-birthweigh t infants and later
neurodevelopment . Acta Paediatr 1996; 85: 1445–9
11. Ellison PH, Horn JL, Browning CA. Constructio n of an infant
neurologica l internationa l battery (INFANIB) for the assessment
of neurologica l integrety in infancy. Phys Ther 1985; 65: 1326–
31
12. Ellison PH. INFANIB. A reliable method for the neuromotor
assessment of infants. Tucson: Therapy Skill Builders; 1994
13. Stavrakas PA, Kemmer-Gacura GE, Engelke SC, Chenier TC.
Predictive validity of the Infant Neurological Internationa l
Battery. Dev Med Child Neurol Suppl 1991; 64: 35–6
14. Bax M. Terminology and classiŽ cation of cerebral palsy. Dev
Med Child Neurol 1964; 6: 295
15. Hagberg B, Hagberg G, Olow I. The changing panorama of
cerebral palsy in Sweden 1954–1970. Acta Pædiatr Scand 1975;
64: 187–92
16. Fagan JF, Shepherd PA. The Fagan test of infant intelligence .
Cleveland, OH: Infanttest Corp; 1987
17. Bailey N. Bailey scales of infant development . New York: The
Psychologica l Corporation; 1969
18. Georgieff MK, Bernbaum JC, Hoffman-Williamso n M, Daft A.
Abnormal truncal muscle tone as a useful early marker for
developmenta l delay in low birth weight infants. Pediatrics 1986;
77: 659–63
19. Stewart KB, Deitz JC, Crowe TK, Robinson N, Bennett FC.
Transient neurologi c signs in infancy and motor outcomes at 41/2
years in children born biologiocall y at risk. Topics Early Child
Spec Educ 1988; 7: 71–83
20. Li AK, Sauve RS, Creighton DE. Early indicators of learning
problems in high risk children. J Dev Behav Pediatr 1990; 11: 1–
6
21. Calame A, Reymond GI, Maherzi M, Roulet M, Marchand C,
Prod’Hom LS. Psychological and neurodevelopmenta l outcome
of high risk newborn infants. Helv Paediatr Acta 1976; 31: 287–
97
22. Drillien CM, Thomson A, Burgoyne K. Low-birthweight
children at early school-age: a longitudina l study. Dev Med
Chil Neurol 1980; 22: 26–47
23. Bylund B, Cervin T, Finnstrøm O, Gaddlin PO, Kernell A, Leijon
I, et al. Morbidity and neurologica l function of very low
birthweight infants from the newborn period to 4 y of age. A
prospectiv e study from the south-eas t region of Sweden. Acta
Paediatr 1998; 87: 758–63
24. Cioni G, Ferrari F, Einspieler Ch, Paolicelli P, Barbani M,
Prechtl H. Comparison between observatio n of spontaneou s
movements and neurologica l examination in preterm infants. J
Peadiatr 1997; 130: 704–11
Received Nov. 16, 2000; revision received April 17 and July 19,
2000; accepted July 20, 2000