ELSEVIER
Prospective Follow-up of Primitive Reflex
Profiles in High-Risk Infants: Clues to an
Early Diagnosis of Cerebral Palsy
D i m i t r i o s I. Z a f e i r i o u , M D , P h D , l o a n n i s G . T s i k o u l a s , M D , P h D , a n d
George M. Kremenopoulos, MD, PhD*
To clarify reflex profiles in the first year of life in
connection with categories of neurologic abnormality,
eight primitive reflexes (i.e., the palmar grasp reflex,
the plantar grasp reflex, the Galant response, the
asymmetric tonic neck reflex, the suprapubic extensor
reflex, the crossed extensor reflex, the Rossolimo re-
flex, and the heel reflex) were prospectively examined
in 204 high-risk infants, of whom 58 developed cere-
bral palsy, 22 had developmental retardation, and 124
were normal at follow-up examination at 2 years of
age. The change in the retention time of reflex activity
for each of these reflexes was characteristic for each
category or type of neurologic abnormality: retention
of palmar grasp reflex, suprapubic extensor reflex,
crossed extensor reflex, Rossolimo reflex, and heel re-
flex in spastic cerebral palsy, as well as retention of
plantar grasp reflex, Galant reflex, and asymmetric
tonic neck reflex in athetoid cerebral palsy and some-
what weaker retention of these reflexes in developmen-
tal retardation (statistical significance P < .001 com-
pared with normally developed patients). These char-
acteristic changes imply that a presumptive diagnosis
can be made in neurologically high-risk infants by ex-
amination of the primitive reflexes, which are of spe-
cific significance among the other neurologic criteria
within the first year of life.
Zafeiriou DI, Tsikoulas IG, Kremenopoulos GM. Pro-
spective follow-up of primitive reflex profiles in high-risk
infants: Clues to an early diagnosis of cerebral palsy. Pe-
diatr Neurol 1995; 13:148-152.
Introduction
The evaluation of posture, muscle tone, and primitive
reflexes is an integral part of the neurologic examination
From the 1st Pediatric Clinic and *Department of Neonatology;
Aristotle University of Thessaloniki; Thessaloniki, Greece.
148 PEDIATRIC NEUROLOGY Vol. 13 No. 2
of infants less than 1 year old [1]. Posture and/or muscle
tone disturbances during the first 12 months of life in
high-risk infants are not always prognostic of the later
diagnosis; some of these infants are normally developed,
while others have various types of cerebral palsy (CP) or
developmental retardation (DR) without motor distur-
bance [2-4]. In such patients, examination of primitive
reflexes could contribute toward a more accurate and
timely diagnosis, as well as toward a differential diagnosis
among neurologic abnormalities [5-7]. In our study, we
examined the primitive reflex profiles of 204 high-risk
infants to clarify their importance in early diagnosis and
differential diagnosis of categories or types of neurologic
abnormalities.
Patients and Methods
All patients were referred to the Department of Developmental Med-
icine of the Ist Pediatric Clinic, Aristotle University of Thessaloniki,
during the first month of life. The study period ranged from January 1993
to March 1995. Two hundred four high-risk infants had a complete
neurodevelopmental examination, including examination of primitive re-
flexes, at age 1, 3, 5, 7, 9, and 11 months, as well as a later follow-up
examination at age 2 years. At 2 years of age the patients were divided
into three categories according to neurologic diagnosis: patients with CP,
patients with DR without motor disturbance, and normally developed
patients.
CP was defined as a nonprogressive disorder of posture and gait due
to a lesion of the immature brain [8]. Clinically, a classification system
according to Ingram [9] and Bax [8], modified by Vojta [6], was used.
According to this classification, spastic diplegia was diagnosed if spas-
ticity in the lower limbs was greater than in the upper limbs: spastic
hemiplegia, if spasticity of the extremities was seen on only one side; and
spastic tetraplegia, if spasticity was more pronounced in the upper than
lower limbs, or at least of the same degree. Athetoid CP was defined as
the presence of athetoid movements with defective regulation of the
muscle tone shifting from hypotonia to normal. A diagnosis of ataxic CP
was made when ataxia was the only prominent symptom, without pyra-
midal or extrapyramidal tract signs.
The CP group comprised 58 patients: 49 patients with a spastic CP (24
with spastic diplegia, 12 with spastic hemiplegia, and 13 with spastic
tetraplegia), 7 patients with athetoid CP, and 2 patients with ataxic CP.
Communications should be addressed to:
Dr. Zafeiriou; Egnatia Street 106; 54622 Thessaloniki, Greece.
Received April 20, 1995; accepted July 5, 1995.
1995 by Elsevier Science Inc. 0887-8994/95/$9.50
SSDI 0887-8994(95)00143-4
Table 1. Eliciting the primitive reflexes*

Reflex Position Method Reaction

Palmar grasp Supine Placing the index finger in the palm of the infant Flexion of fingers--fist making
Plantar grasp Supine Pressing a thumb against the sole just behind the Flexion of toes
toes in the foot
Galant Prone Scratching the skin of the infant's back from the Incurvation of the trunk, with the concavity on the
shoulder downwards, 2-3 cm lateral to the stimulated side
spinous processes
Asymmetric tonic neck Supine Rotation of the infant's head to one side for 15 s Extension of the extremities on the chin side and
reflex flexion of those on the occipital side
Suprapubic extensor Supine Pressing the skin over the pubic bone with the Reflex extension of both lower extremities, with
fingers adduction and internal rotation into talipes
equinus
Crossed extensor Supine Passive total flexion of one lower extremity Extension of the other lower limb with adduction
and internal rotation into talipes equinus
Rossolimo Supine Light tapping of the 2nd-4th toes at their plantar Tonic flexion of the toes at the 1st
surface metacarpophalangeal joint
Heel Supine Tapping on the heel with a hammer, with the hip Rapid reflex extension of the lower extremity in
and knee joint flexed, and the ankle joint in a question
neutral position

* Based on data from References 6 and 11.

The CP patients were further classified as of prenatal, perinatal, or un-
known etiology according to Hagberg [10]. In 25 (43.1%) patients CP
was considered to be of prenatal etiology and in 21 (36.2%) patients of
perinatal etiology; in the remaining 12 (20.7%) patients no clear etiology
for the CP could be found.
DR was diagnosed as a developmental quotient (DQ) <75 in Griffiths
Mental Developmental Scale at the age of 15 months. This group con-
sisted of 22 patients, 3 of whom had Down syndrome, 1 had congenital
hydrocephalus, 3 had intraventricular hemorrhage (2 grade II and 1
grade Ill), 3 had neonatal convulsions, and 12 had no clear etiology.
The ratio of preterm to term infants was 32/26, 4/18, and 54/70 in the
CP group, the DR group, and the normal group, respectively. For pre-
term infants, the age at the time of each neurodevelopmental examination
was expressed as the corrected age according to the expected birth date.
The primitive reflexes examined in our study were the palmar grasp
reflex, the plantar grasp reflex, the Galant reflex, the asymmetric tonic
neck reflex (ATNR), the suprapubic extensor reflex, the crossed extensor
reflex, the Rossolimo reflex, and the heel reflex. Their selection from
among other neurologic signs or criteria was made because the judge-
ment of their reactions is simple and their clinical significance has been
suggested in the literature [6,7,11 ]. The manner of eliciting each of these
primitive reflexes and the expected reactions are shown in Table 1.
All neurodevelopmental examinations were performed by one of the
authors (DIZ), with the patient lying quietly on the examination bed in
the waking state. Each reflex was obtained 3 or 4 times for evaluation of
the response, and the reaction was classified as positive only in those
cases where the mean of the positive responses was >0.5 of all trials.
The results were statistically analyzed by means of chi-square analysis
and Fisher's Exact Test to evaluate the differences in reflex reactivity
between normally developed patients and those with various neurologic
abnormalities at each age of examination.
Results
The palmar grasp reflex (Table 2) was preserved in the
majority of normal infants until age 5 months, after which
it began to disappear. Most spastic patients exhibited full
activity of the reflex until age 11 months (statistical sig-
nificance P < .001 compared with normal infants). The
reflex profile in athetoid, ataxic, and DR patients was not
different from normal.
The plantar grasp response (Table 2) exhibited nearly
full activity until 7 months of age in the normal infants,
while it could be elicited in only 1/5 of the patients at age

Table 2. Retention time of palmar grasp reflex and plantar grasp reflex

Examination Spastic CP Athetoid CP Ataxic CP DR Normal

age (n) (n) (n) (n) (n)
(mos) PaR PrR PaR PrR PaR PrR PaR PrR PaR PrR

1 49 49 7 7 2 2 22 22 119 124
3 49 37* 6 7 2 2 22 22 119 121
5 48 28* 6 7 2 2 21 22 119 117
7 44* 20* 4 7 1 2 9 20 24 93
9 41" 15" 1 7 1 1 5 14" 8 70
11 41" 9* 1 7* -- 1 4 12" 5 25

* Statistical significance: P < .001 compared with normal patients.

Abbreviations:
CP = Cerebral palsy n = Number of patients with positive reflex PrR = Plantar grasp reflex
DR = Developmental retardation PaR = Palmar grasp reflex

Zafeiriou et al: Prospective Follow-up of Primitive Reflex 149

Table 3. Retention time of Galant reflex and asymmetric tonic neck reflex

Examination Spastic CP Athetoid CP Ataxic CP DR Normal

age (n) (n) (n) (n) (n)
(mos) GR ATNR GR ATNR GR ATNR GR ATNR GR ATNR

1 49 49 7 7 2 2 22 22 124 124
3 44 45* 7 6 2 2 18 18 102 58
5 39* 39* 7* 6* l I 17" 16' 10 11
7 34* 39* 7* 6* -- -- 14' 10" 8 --
9 24* 33* 7* 5* -- -- 5* 4* 5 -
11 8* 28* 7* 5* -- -- 4* 2 4 --

* Statistical significance: P < .001 compared with normal patients.

Abbreviations:
A T N R = A s y m m e t r i c tonic neck reflex DR = Developmental retardation n : N u m b e r of patients with positive reflex
CP = Cerebral palsy GR = Galant reflex

11 months. In patients with athetoid CP, the response was
strongly retained until age 11 months (P < .001), while in
spastic patients, there was a gradual decrease of reflex
reactivity from 3 to 11 months of age (P < .001). DR
infants displayed a longer retention of the reflex compared
to normal patients, which was statistically significant at
the age of 9 months and beyond.
In most of the normal infants, the Galant reflex (Table
3) started to disappear at age 3 months and was completely
gone by age 9 months. In all athetoid CP patients, Galant
reflex remained positive until age 11 months (P < .001
compared to normal infants />5 months old), while in
spastic CP patients the Galant reflex disappeared more
slowly than in normal patients, remaining positive in
about 2/5 of them at age 9 months. Infants with DR also
manifested a prolonged retention of Galant reflex which
was statistically significant from age 5 months (P < . 001).
In about half of the normal infants, ATNR (Table 3),
started to disappear at age 5 months, with total disappear-
ance by age 7 months. In spastic or athetoid infants,
ATNR was retained for a longer period than in normal
infants, the difference in reflex reactivity being statisti-
cally significant (P < .001) at age 3 months for the spastic
group and at age 7 months for the athetoid group. In the 2
patients with ataxic CP, ATNR disappeared by age 7
months, while children with DR manifested prolonged re-
tention of the above reflex, which was statistically signif-
icant from 3 to 9 months of age (P < .001).
Similarities in the reflex profiles of all patient groups
were observed for the following primitive reflexes: supra-
pubic extensor reflex, crossed extensor reflex, Rossolimo
reflex, and heel reflex (Tables 4, 5). In patients with nor-
mal development, these primitive reflexes were strongly
retained until the first month of age and then rapidly de-
creased, disappearing by age 5 months. In infants with
spastic CP, these reflexes were retained much longer than
in normal infants, to 11 months of age (statistical signif-
icance, P < .001). Infants with athetoid CP, ataxic CP,
and DR displayed reflex profiles similar to those of normal
infants.
The above results are summarized in Table 6. There are
clear differences in the reflex activity of the palmar grasp
reflex, suprapubic extensor reflex, crossed extensor re-
flex, Rossolimo reflex, and heel reflex between infants
with spastic and athetoid CP. The 2 patients with ataxic
CP manifested no differences in the retention time of these
primitive reflexes compared with normal patients; how-
ever, their number is too small to allow any statistical

Table 4. Retention time of suprapubic extensor reflex and crossed extensor reflex

Examination Spastic CP Athetoid CP Ataxic CP DR Normal

age (n) (n) (n) (n) (n)
(mos) SER CER SER CER SER CER SER CER SER CER

1 49 49 7 7 2 2 22 19 124 105
3 45* 48* 3 2 1 -- 6 5 102 19
5 43* 44* 1 -- -- -- 3 2 10 3
7 41" 41' . . . . 1 -- 8 ---
9 41" 41" . . . . 1 -- 5 --
11 40* 39* . . . . 1 -- 4 --

* Statistical significance: P < .001 compared with normal patients.

Abbreviations:
CER = Crossed extensor reflex DR = Developmental retardation SER = Suprapubic extensor reflex
CP = Cerebral palsy n = N u m b e r of patients with positive reflex

150 PEDIATRIC NEUROLOGY Vol. 13 No. 2

Table 5. Retention tirae of Rossolimo reflex and heel reflex

Examination Spastic CP Athetoid CP Ataxic CP DR Normal

age (n) (n) (n) (n) (n)
(mos) RR HR RR HR RR FIR RR HR RR FIR

1 48 45 7 7 2 2 20 22 102 115
3 46* 42* 1 2 -- 1 4 7 15 25
5 45* 40* -- 1 -- -- -- 5 -- 11
7 45* 39* . . . . . 2 -- 3
9 45* 39* . . . . . . . .
11 45* 38* . . . . . . . .

* Statistical significance:P < .001 compared with normal patients.

Abbreviations:
CP = Cerebralpalsy HR = Heel reflex RR = Rossolimoreflex
DR = Developmentalretardation n = Numberof patients with positive reflex

conclusions. DR patients demonstrated the same tendency
as athetoid patients for retention of plantar grasp reflex,
Galant reflex, and ATNR, although this tendency was
stronger in the athetoid group.
Discussion
In the present prospective study, we analyzed the prim-
itive reflex profiles in high-risk infants referred for neu-
rodevelopmental examination in the Department of Devel-
opmental Medicine during the first year of life. Our results
concerning the reflex profiles for normal infants are con-
sistent with other studies which used the same method of
eliciting reflexes [6,11,12], thus indicating the validity of
the method used. From our results, we have concluded
that early diagnostic clues as to whether a high-risk infant
will become normal or abnormal at the age of 2 years is
the presence of the suprapubic extensor reflex, the crossed
extensor reflex, the Rossolimo reflex, and the heel reflex
after 3 months; a positive Galant reflex and ATNR reflex
after 5 months; a positive palmar grasp reflex after 7
Table 6. Retention of reflex activity in infants with
neurologic abnormalities
months; and a negative plantar grasp reflex after 3 months.
These findings indicate a poor prognosis; on the basis of
these primitive reflexes, neurologically abnormal children
could be in part diagnosed earlier (after 3 months of age).
Examination of these eight primitive reflexes (Table 6)
can clearly distinguish between spastic and athetoid CP.
The diagnosis of ataxic CP, as well as the distinction
between athetoid CP and DR, is not possible with exam-
ination of the primitive reflexes alone. The combination of
posture, muscle tone, and primitive reflex examination
can help in such presumptive diagnoses [ 13,14].
The neurodevelopmental evaluation of an infant should
therefore be comprehensive, including observation of pos-
ture, voluntary and involuntary movements, muscle tone,
primitive reflexes, as well as other responses and signs.
The neurologic status of an infant cannot be judged on the
basis of a single criterion, because there exists no criterion
that predicts the prognosis in all cases [15]. On the other
hand, the examiner should critically review the practical
significance of each procedure, so as to more easily reach
an early diagnosis. Our study indicates that examination of
the eight primitive reflexes beginning at age 3 months is a
useful diagnostic tool for early diagnosis of CP.

Spastic Athetoid Ataxic References

Reflex CP CP CP DR
[1] MenkesJH. Neurologicalexaminationof the child and infant.
Palmar grasp + 0 0 0 In: Menkes JH, ed. Textbook of child neurology. 4th ed. Philadelphia:
Plantar grasp - ++ 0 + Lea & Febiger, 1990:12-16.
Galant + ++ 0 + [2] Piper MC, Mazer B, Silver KM, Ramsay M. Resolution of
Asymmetric tonic neck neurologicalsymptomsin high-riskinfants during the first two years of
reflex + ++ 0 + life. Dev Med Child Neurol 1988;30:26-35.
Suprapubic extensor + 0 0 0 [3] Burns YR, O'Callaghan M, Tudehope DI. Early identification
Crossed extensor + 0 0 0 of cerebral palsy in high risk infants. Aust Paediatr J 1989;25:215-9.
Rossolimo + 0 0 0 [4] PeBenito R, Santello MD, Faxas TA, Ferretti C, Fisch CB.
Heel + 0 0 0
Residual developmentaldisabilities in children with transient hyperto-
nicity in infancy. Pediatr Neurol 1989;5:154-60.
Abbreviations:
CP = Cerebralpalsy [5] Capute AJ. Identifyingcerebral palsy in infancy through study
DR = Developmentalretardation of primitivereflex profiles. Pediatr Ann 1979;8:34-42.
+ + = Markedlyretained reflexes [6] Vojta V. Die cerebralen Bewegungstoerungenim Kindesalter.
+ = Slightly retained reflexes 4. Auflage. Stuttgart: FerdinandEnke Vedag, 1988.
0 = Unchangedreflexes [7] Futagi Y, Tagawa T, Otani K. Primitive reflex profiles in in-
- = Decreased reflexes fants: Differences based on categories of neurological abnormalities.
Brain Dev 1992;14:294-8.

Zafeiriou et al: Prospective Follow-upof PrimitiveReflex 151

 [8] Bax CO. Terminology and classification of cerebral palsy. Dev
Med Child Neurol 1964;6:295-7.
[9] hlgram "ITS. Paediatric aspects of cerebral palsy. Edinburgh: E
& S Livingstone, 1964.
[10] Hagberg B, Hagberg G, Olow I, yon Wendt L. The changing
panorama of cerebral palsy in Sweden, V. The birth year period 1979-
1982. Acta Paediatr Scand 1989;78:283-90.
[11] Towen B. Neurological development in infancy. Philadelphia:
JB Lippincott, 1976.
[12] Paine RS, Brazelton TB, Donovan DE, Drorbaugh JE, Hubbell
152 PEDIATRIC NEUROLOGY Vol. 13 No. 2
JP, Sears EM. Evolution of postural reflexes in normal infants and in the
presence of chronic brain syndromes. Neurology 1964;4:1036-48.
[13] Soiomons G, Holden RH, Denhoff E. The changing picture of
cerebral dysfunction in early childhood. J Pediatr 1963;63:113-20.
[14] Taudorf K, Hansen FJ, Melchior JC. Spontaneous remission of
cerebral palsy. Neuropediatrics 1986;17:19-22.
[15] Miller G. The cerebral palsies. In: Miller G, Ramer JC, eds.
Static encephalopathies of infancy and childhood. New York: Raven
Press, 1992:11-26.