Early feeding
with cerebral palsy (CP). Previous studies (Shapiro et al.
problems in children
with cerebral palsy:
weight and neuro-
developmental
outcomes
Sara Motion MRCP, Specialist Paediatric Registrar;
Kate Northstone MSc, Research Statistician;
Alan Emond* FRCP, Consultant Paediatrician, Unit of
Paediatric and Perinatal Epidemiology, Institute of Child
Health, Bristol;
Sally Stucke MRCP, Professor of Community Child Health,
Neville Hall Hospital, Abergevenney;
Jean Golding PhD, Professor of Paediatric and Perinatal
Epidemiology and the Avon Longitudinal Study of Parents
and Children Study Team, Unit of Paediatric and Perinatal
Epidemiology, Institute of Child Health, Bristol, UK.
*Correspondence to third author at Unit of Paediatric and
Perinatal Epidemiology, Institute of Child Health, 24 Tyndall
Avenue, Bristol BS8 1TQ, UK.
E-mail: alan.emond@bristol.ac.uk
Using prospectively collected data from 13 971 births enrolled
in a large population-based cohort study (Avon Longitudinal
Study of Parents and Children: ALSPAC), the prevalence of
feeding difficulties at 4 weeks and 6 months of age in 33
children subsequently diagnosed with cerebral palsy (CP)
were investigated. It was also assessed whether early feeding
difficulties could be predictive of functional and growth
outcomes at age 8 years. Weak sucking at 4 weeks of age was
reported in 11 of 23 children with CP and in 2206 of 12 299
(18%) of the remaining population (p<0.0001), and great
difficulties feeding at 6 months of age was reported in two of
21 (10%) of the group with CP and in 373 of 10 941 (3.3%) of
control participants (p=0.017). Feeding difficulties at 4
weeks of age were associated with the pattern of functional
impairment at age 4 years (p=0.009) and at 8 years
(p=0.068), being clinically underweight (p=0.01), and
having speech and swallowing difficulties (p=0.005) at 8
years of age. Early, persistent, and severe feeding difficulties
are a marker for subsequent poor growth, feeding, and
developmental outcomes and can identify children with CP
who will benefit from gastrostomy feeding.
40 Developmental Medicine & Child Neurology 2002, 44: 40–43
Feeding problems are common in all age groups of children
1986, Boyle 1991, Reilly et al. 1996) have reported a preva-
lence of feeding problems in the first year of life of between
38% and 57%, using retrospectively collected data, on chil-
dren followed-up once CP had been diagnosed. In a signifi-
cant number of children, these feeding difficulties preceded
the diagnosis. Most research has reported on case series iden-
tified from child development centres and special interest
clinics and, therefore, may suffer from ascertainment bias.
Feeding skills emerge in a predictable sequence in normal-
ly developing children, and the progression of these skills is
usually associated with the sequential attainment of motor,
language, and social milestones (Bartz and Deubler 1990).
Feeding is a complex task requiring significant neuromuscu-
lar coordination, and integration of motor and sensory path-
ways. Neuromotor and neurosensory dysfunction affecting
the oropharynx and oesophagus is likely to be behind the
early feeding difficulties in children who will subsequently be
diagnosed with CP (Shapiro et al. 1986).
There is little population-based data on the prevalence of
feeding problems in the normally developing population,
and there is a paucity of prospectively collected data on
infant feeding difficulties in CP. The relation of early feeding
problems in CP with subsequent growth and neurodevelop-
mental outcomes appears not to have been studied before.
The aims of this study were to: (1) identify and characterize
a representative CP cohort, derived from a longitudinal popu-
lation-based study; (2) determine the frequency of feeding
problems in the normally developing population and the CP
cohort; (3) determine whether feeding problems at 4 weeks of
age were predictive of a subsequent diagnosis of CP; and (4)
explore the relation in children with CP between infant feed-
ing problems and weight, and neurodevelopmental outcomes
at 4 and 8 years of age
Method
All births to women resident in the former Avon Health
Authority area, with an expected date of delivery between 1
April 1991 and 31 December 1992 were eligible for enrol-
ment in the Avon Longitudinal Study of Parents and Children
(ALSPAC), a population-based study investigating environ-
mental and genetic influences on the health and development
of children (see ALSPAC website: www.ich.bristol.ac.uk). Over
80% of the known births from the geographically defined
catchment area were included, resulting in a total cohort of
13 971 surviving children. This study population has social
and demographic characteristics in common with national
census surveys (Golding 1996).
From this birth cohort, every effort was made to ensure total
case ascertainment at 4 to 5 years of age of children diagnosed
with CP, using the following sources: (1) the ALSPAC self-com-
pletion questionnaires in which the parent documented any
diagnoses made on their children; (2) child health records
(using ICD coding for CP; World Health Organization 1989);
(3) the local hospital patient administration system (PAS; using
ICD codings for CP for hospital admissions); (4) a formal
inquiry to each of the consultant paediatricians responsible for
the child development centres located within the study area,
to identify children with a possible diagnosis of CP and with
birth dates eligible for this study.
This initial search produced a short list of 51 individuals,
from a birth cohort of 13 971 live births arising from a total tion, with Fisher’s exact test used where appropriate. Analysis
population of 1 million. The diagnoses were reviewed in was performed using SPSS software (version 9.0). Ethical per-
1996, when the children were approximately 4 years of age, mission for the ALSPAC study was granted by the ethics com-
by a consultant paediatrician who examined each child (and mittees of United Bristol Healthcare Trust, Frenchay and
in a small minority of cases where this was not possible, the Southmead Healthcare Trusts, and the study is also monitored
medical records were examined). by the ALSPAC Ethics and Law Advisory Committee.
The CP study group was then classified using the Standard
Recording of Central Motor Deficit form (Evans et al. 1989), Results
with information recorded on the pattern of CP, the function- CHARACTERISTICS OF CHILDREN DIAGNOSED WITH CP
al motor impairment, any oromotor dysfunction, and com- Thirty-seven children with a diagnosis of CP were identified
munication impairment. from a birth cohort of 13 971, giving a point prevalence at age
At 7 to 8 years of age (school age), a questionnaire was com- 4 years of 2.6 per 1000. Three children moved away from the
pleted on these children by the school doctor (in a small num- study area, and for these the outcome data are incomplete.
ber of cases the school nurse), together with a review of the One child died at the age of 5 years 6 months from respiratory
medical records. The established pattern of CP, the extent of complications. For outcome data, therefore, the study group
the oromotor dysfunction (swallowing difficulties, texture and comprises 33 individuals. Due to inconsistency in the num-
method of feeding, speech impairment, control of saliva), and ber of questionnaires completed, the denominator varies for
the weight of each child were recorded. many of the outcomes.
Weight records of each child with CP were examined to Table I outlines the characteristics of the CP group at birth.
determine whether there was significant faltering of growth. Over half of the children were born at term, with weights of
Small for gestational age (SGA) participants were defined as over 2.5 kg, and 4 out of 37 of the group were SGA at birth.
those having a birthweight, corrected for gestation, below the
10th centile using UK reference growth charts (Freeman et al. PREVALENCE OF FEEDING PROBLEMS
1995). At 9 and 18 months of age the weights recorded by Parental reporting of weak sucking at 4 weeks of age, and
health visitors at standardized developmental contacts were extreme difficulty in feeding at 4 weeks and 6 months of age,
used, and at 7 to 8 years of age the weights used were those doc-
umented at annual school health assessments. From these data
the z-scores (weight standard deviation score) were derived,
and a child was defined as being clinically underweight with a Table I: Clinical characteristics of group with CP
single weight z-score of ≤ –2.4 (equivalent to the 0.4 centile on Clinical characteristic CP group (n=37)
the UK reference growth chart; Freeman et al. 1995).
Parents documented feeding behaviours of the children Gestation
using self-completion postal questionnaires sent out at the >37 wk 19/37
child’s chronological (uncorrected) age. Parents were asked 32–37 wk 9/37
to complete the questionnaire even if their child was still in the <32 wk 9/37
special care baby unit. Questions on feeding behaviour were Birthweight
selected from several clinical and research questionnaires, >2500g 20/37
1501–2500g 9/37
and piloted on the Avon population before being used in the
≤ 1500g 8/37
ALSPAC study. Parents were asked to rate on a scale of 1 to 5 SGA at birth 4/37
whether their baby demonstrated specific feeding behaviours, Sex ratio M:F 22:15
which were named but not described in the questionnaire. No Neurological pattern at 4 years
definitions were provided on what constituted a ‘problem’. 4-limb CP 15/34a
The responses used were those to specific questions on feed- 2-limb CP 19/34
ing contained in questionnaires completed at 4 weeks and 6 Neurological pattern at 8 years
months of age, which would have preceded the diagnosis of 4-limb CP 16/33
CP in most of the cases. 2-limb CP 17/33
Analysis using χ2 was performed to detect any differences in
a Variation in denominator reflects the number of available replies.
the proportions of those with CP and the remaining popula-

Table II: Feeding problems in group with CP

Clinical characteristic CP study groupa Rest of populationa % p

(n=37) (n=12 322)
At 4 wks
Weak suck 11/23 2206/12299 18 <0.001
Extremely difficult to feed 3/23 115/12218 1 <0.001
Exhaustion with feeding 13/23 6086/12322 55 0.351
Choking with feeding 10/23 6832/12322 55 0.346
At 6 mo
Great difficulties with feeding 2/21 373/10941 3 0.017
a Variation in denominator reflects the number of parental responses.

Feeding Problems in Cerebral Palsy Sara Motion et al. 41

were found to be more frequent in the CP cohort compared
with the remaining population (both p<0.001). Exhaustion
and choking with feeding at 4 weeks were frequently report-
ed in the CP study group, but were also reported by the rest
of the population (Table II).
Oromotor dysfunction was less frequently described in the
CP group at 8 years of age, with improvements in swallowing
(six of 33 reporting difficulties at 8 years of age, compared with
12 of 26 at 4 years), and there was a similar improvement in the
control of saliva. The proportion of those who were under-
weight increased from eight of 33 at 4 years to 10 of 33 at age 8
years (Table III). These trends were not statistically significant.
RELATION BETWEEN EARLY FEEDING DIFFICULTIES AND OUTCOME
AT 4 AND 8 YEARS
There was an association between weak sucking and exhaus-
tion with feeding at 4 weeks of age, and severe functional
impairment of the upper limb and lower limb at 4 years of age.
Nine of 10 children with CP reported to have weak sucking at 4
weeks of age were unable to walk at 4 years, compared with
three of eight children who were reported to suck normally at 4
weeks of age (p=0.009). Of the 11 children with CP who were
reported to experience exhaustion with feeding at 4 weeks, 10
were unable to assist with dressing and feeding at 4 years, com-
pared with one of six of the participants with CP who were not
reported to feel exhaustion with feeds (p=0.009).
The other feeding difficulties (choking and great difficul-
ties feeding at 4 weeks, and poor sucking at 6 months) did
not show a significant correlation with neurological impair-
ment in the preschool age group.
At 8 years of age the diagnosis of four-limb CP was present
in seven of 10 children whose parents had reported exhaus-
tion with feeding at 4 weeks of age, compared with one of six
children who were not reported to be exhausted by feeding
at 4 weeks (p=0.068). There were no significant associations
between the other 4-week feeding behaviours and the subse-
quent diagnosis of four-limb CP in school-age children. The
children diagnosed with two-limb CP were less likely to have
reported feeding difficulties at 4 weeks (p=0.05).
Five of the 11 children with CP who had persistent dribbling
of saliva at the age of 8 years had been reported to show exhaus-
tion with feeding at 4 weeks of age, compared with five of 21
children who did not have persistent dribbling (p=0.005). Of
the nine children who still required food of modified texture at
8 years of age, five had been reported to show exhaustion with
feeding at 4 weeks, compared with five of 23 who were able to
take food of a normal texture (p=0.05). A trend was also
observed between swallowing and speech impairment at the
age of 8 years, and exhaustion with feeding at 4 weeks of age
Table III: Outcome measures in the group with CP
Clinical characteristic Age 4 y (n=34)a Age 8 y (n=33)
Oromotor
Swallowing difficulties 12/26 6/33
Dribbling of saliva 14/29 11/33
Speech impairment 14/28 15/33
Growth
Underweight 8/33 10/33
a Variation in denominator reflects the number of parental responses.
42 Developmental Medicine & Child Neurology 2002, 44: 40–43
(p=0.068). There was no statistically significant association
between the other difficult feeding behaviours at 4 weeks and
the oromotor dysfunction at school age.
Five of 10 children who were underweight at 8 years of
age were reported to have exhaustion with feeding at 4
weeks, compared with five of 22 who were not underweight
(p=0.01). Of those with feeding problems at 6 months, three
of four were underweight at 8 years, compared with one of
four in those who did not have early feeding problems
(p=0.013). Looking at the progression of the weight charac-
teristics for the child with CP as they grow older, small size at
birth was documented in four of 37 infants, at 4 years of age
eight of 34 of the children were underweight, and by 8 years
of age the proportion increased to 10 of 33 children. Of the
10 children who were underweight at 8 years, nine had a
four-limb pattern of CP. Only one child (one of 33) was being
fed via a gastrostomy.
Discussion
Feeding difficulties reported by parents in the first 4 weeks of
life in a cohort of children diagnosed with CP have been
shown to be associated with more severe neurodevelopmen-
tal impairments and increased likelihood of being under-
weight by school age in this study. The recognition that infant
feeding difficulties may indicate a more severely affected child
can help professionals both to highlight the children who may
benefit from assessments and interventions, such as gastros-
tomies (demonstrated to be associated with improved long-
term growth and quality of life outcomes in CP, e.g. Smith et al.
1999), and also potentially to assist in counselling families
regarding the long-term prognosis for their affected child.
The strengths of this study lie in the longitudinal, popula-
tion-based data which give a clear picture of the prevalence
of feeding difficulties, both in the general population and in
a cohort of children with CP (in whom data collection was
possible in a prospective and unbiased way). The main weak-
ness intrinsic to the methodology was the lack of definition
in the parental questionnaires of what constituted feeding
difficulties, particularly ‘choking’ and ‘exhaustion’. We had
expected feeding problems at 4 weeks of age to be common-
ly reported in those with CP, but we had not anticipated the
frequency of difficult feeding behaviours described by par-
ents with young infants in the rest of the population. The rea-
sons why so many parents reported difficult feeding patterns
at this age are numerous, with a complex interrelation of
maternal and infant factors, possibly, in new and sometimes
underconfident mothers who may not have formed secure
attachments with their young infants. The wide reporting of
symptoms such as choking and exhaustion with feeding at
this age, together with the relatively small study group with
CP, meant that infant feeding difficulties by themselves had
poor predictive value in identifying children who would sub-
sequently be diagnosed with CP.
A second difficulty with the dataset is that a proportion of
parents did not return all the questionnaires for a variety of
reasons, perhaps in some cases due to anxiety created by the
child’s medical problems or delays in their developmental
progression. With respect to the infants born at less than 32
weeks’ gestation, seven of nine of these children were them-
selves the participants in a separate study (Avon Pre-term
Infant Project Team 1998), which was probably a source of
confusion to some families. These factors have resulted in
considerable variability in the availability of questionnaire
data for some of the outcomes.
The assessment of feeding abilities in infants of 4 weeks of
chronological age may seem early for this CP group, among
whom almost half were delivered preterm. However, we
were limited by the timing of the questionnaires distributed
in the ALSPAC study, which in the first 6 months of life were at
4 weeks and 6 months of age only. Prematurity itself was a sig-
nificant confounding variable for feeding difficulties both in
the group with CP and in the rest of the population, as we
have reported elsewhere (Motion et al. 2001)
In clinical practice, professionals traditionally pay attention
to the reporting of weak sucking and choking with feeding in
infants suspected of having problems with either growth or
development. In this study, however, we have found that these
symptoms were too frequently reported by parents to be use-
ful, and that exhaustion with feeding at 4 weeks of age, and
great difficulties feeding at age 6 months were the two most
useful symptoms for identifying the severely affected child
with CP: these factors were significantly associated with four-
limb CP, severe oromotor dysfunction, and being underweight
at school age. The persistence and the severity of the feeding
symptoms seem to be the most important predictive factors. A
detailed feeding history in early infancy can thus be helpful in
identifying children who are likely to have relatively poor out-
come at school age.
Other authors have described the assessment and manage-
ment of oromotor dysfunction in the child with CP (Gisel and
Patrick 1988, Boyle 1991, Reilly and Skuse 1992, Wolke and
Skuse 1992) and our study clearly highlights the persistence of
this problem in the school-aged child with four-limb CP. A sig-
nificant proportion of the CP cohort with severe oromotor dif-
ficulties were underweight at their school health assessments,
and these observations confirm the prevalence of relative
undernutrition in this type of CP. In fact, although the numbers
of cases were small, there appeared to be less oromotor dys-
function at 8 years compared with 4 years of age, while there
was an increase in the proportion of older children who were
underweight. The possible explanations for this are either that
feeding problems at 8 years were being overlooked, or because
of an increase in metabolic demands at 8 years due to increased
energy consumption. However, whether due to poor calorie
intake or increasing energy expenditure, or a combination of
the two, it is clear that the nutritional needs of this group of
children with CP had not been adequately met: one obvious
implication of this finding is that we may not be using enough
gastrostomies for feeding. Although there is a growing body
of evidence regarding the benefits of gastrostomies on out-
comes in CP such as weight and family dynamics (Eltuni and
Sullivan 1997), professionals and families continue to be
reluctant to use this feeding method to prevent undernutri-
tion. We anticipate that weight outcomes could be improved
by the earlier use of gastrostomy feeding, particularly in the
child with four-limb CP with early and continuing feeding dif-
ficulties. Conversely, we can also conclude that a diagnosis of
two-limb CP is likely to be associated with normal growth and
oromotor function in the majority of cases, and gastrostomy
insertion would not be indicated for this group.
In conclusion, we have identified that feeding problems
in infants at 4 weeks of age are commonly reported by par-
ents of children who will subsequently be diagnosed with CP.
However, these are also frequently reported by the rest of the
population. Exhaustion with feeding at age 4 weeks and per-
sisting feeding difficulties at 6 months are associated with
more severe neurodevelopmental and growth impairments
in school age children with CP. Assessing these difficult feed-
ing behaviours may help to identify a group of children who
may benefit from early assessments, monitoring, and inter-
ventions including gastrostomy insertion.
Accepted for publication 24th August 2001.
Acknowledgements
We are extremely grateful to all the parents who have taken part in this
study for several years and continue to do so, and the midwives for
their cooperation and help in recruiting the mothers during
pregnancy. The whole ALSPAC study team comprising interviewers,
computer technicians, clerical workers, research scientists,
volunteers, and managers continue to make the study possible. The
ALSPAC study could not have been undertaken without the financial
support of the Medical Research Council, the Wellcome Trust, the
Department of Health, the Department of the Environment, and the
Ministry of Agriculture, Fisheries and Food. The ALSPAC study is part
of the World Health Organization initiated European Longitudinal
Study of Pregnancy and Childhood.
References
Avon Pre-term Infant Project Team. (1998) A randomised trial of
parental support for families with very pre-term children.
Archives of Disease in Childhood 79: F4–11
Bartz AH, Deubler DC. (1990) Identification of feeding and
nutrition problems in young children with neuromotor
involvement: a self-assessment. Journal of Pediatric and
Perinatal Nutrition 2: 1–15.
Boyle JT. (1991) Nutritional management of the developmentally
disabled child. Pediatric Surgery International 6: 76–81.
Eltuni M, Sullivan PB. (1997) Nutritional management of the
disabled child: the role of percutaneous endoscopic gastrostomy.
Developmental Medicine & Child Neurology 39: 66–8.
Evans P, Johnson A, Mutch L, Alberman E. (1989) A standard form
for recording clinical findings in children with a motor deficit of
central origin. Developmental Medicine & Child Neurology
31: 119–27.
Freeman JV, Cole TJ, Chinn S, Jones PRN, White EM, Preece MA.
(1995) Cross-sectional stature and weight reference curves for
the UK, 1990. Archives of Disease in Childhood 73: 17–24.
Gisel EG, Patrick J. (1988) Identification of children with cerebral palsy
unable to maintain a normal nutritional state. Lancet 1: 283–6.
Golding J, ALSPAC Study Team. (1996) Children of the 90’s: a
resource for assessing the magnitude of long-term effects of pre-
natal and perinatal events. Contemporary Reviews in Obstetrics
and Gynaecology 8: 89–92.
Motion S, Emond A, Northstone K, ALSPAC Study Team. (2001)
Persistent early feeding difficulties and subsequent growth and
developmental outcomes. (submitted for publication).
Reilly S, Skuse D. (1992) Characteristics and management of
feeding problems of young children with cerebral palsy.
Developmental Medicine & Child Neurology 34: 379–88.
— — Poblete X. (1996) Prevalence of feeding problems and oral
motor dysfunction in children with cerebral palsy: a community
survey. Journal of Pediatrics 129: 877–82.
Shapiro BK, Green P, Krick J, Allen D, Capute AJ. (1986) Growth of
severely impaired children: neurological versus nutritional
factors. Developmental Medicine & Child Neurology 28: 729–33.
Smith SW, Camfield C, Camfield P. (1999) Living with cerebral palsy
and tube feeding: a population-based follow-up study. Journal of
Pediatrics 135: 307–10.
Wolke D, Skuse D. (1992) The management of infant feeding
problems. In: Cooper P, Stein A (eds). The Nature and
Management of Feeding Problems and Eating Disorders in Young
People. New York: Harwood Academic Publications. p 29–59.
World Health Organization (1989). International Classification of
Diseases, Ninth Revision. Geneva: World Health Organization.
Feeding Problems in Cerebral Palsy Sara Motion et al. 43