Impact of feeding problems on nutritional intake and growth: Oxford Feeding Study II

P B Sullivan* MA MD FRCP FRCPCH, University Lecturer in Paediatrics, Department of Paediatrics; E Juszczak, Statistician, Centre for Statistics in Medicine; B R Lambert SRD MSc, Paediatric Dietician; M Rose MRCP MRCPCH, University of Oxford; M E Ford-Adams MRCP MRCPCH, Paediatric Registrar, Department of Paediatrics, John Radcliffe Hospital; A Johnson MA FRCP FRCPCH, National Perinatal Epidemiology Unit, University of Oxford, Oxford, UK. *Correspondence to rst author at University of Oxford, Department of Paediatrics, John Radcliffe Hospital, Oxford, OX3 9DU, UK. E-mail: peter.sullivan@paediatrics.ox.ac.uk

Poor nutritional status and growth failure are common in children with cerebral palsy (CP). The aim of this study was to assess, within a subgroup of a large and clearly dened population of children with disabilities, the impact of feeding difculties on (1) the quality (micronutrient intake) and quantity (macronutrient intake) of their diet and (2) their growth. One hundred children with disabilities (40 females, 60 males; mean age 9 years, SD 2 years 5 months; range 4 years 6 months to 13 years 7 months) underwent a detailed dietetic analysis and a comprehensive anthropometric assessment. Diagnostic categories of disability were: CP (n=90); global developmental delay (n=3); Marfan syndrome (n=1); intractable epilepsy (n=2); agenesis of the corpus callosum (n=2); methyl malonic aciduria (n=1); and congenital rubella (n=1). Neurological impairment was classied according to difculty with mobility which was graded as mild (little or no difculty walking), moderate (difculty walking but does not need aids or a helper), and severe (needs aids and/or a helper or cannot walk). Results conrmed the signicant impact of neurological impairment in children on body growth and nutritional status becoming worse in those with a greater degree of motor impairment. The major nutritional decit was in energy intake, with only one fth reportedly regularly achieving over 100% estimated average requirement (EAR), whilst micronutrient intake was less markedly impaired and protein intake was normal in this group (96% above EAR). Many children with neurological impairment would benet from individual nutritional assessment and management as part of their overall care.

Poor nutritional status and growth failure are common in children with cerebral palsy (Thommessen et al. 1991, Stallings et al. 1995). It has been reported in several studies that such growth failure can be related to an inadequate food intake, resulting from self-feeding impairment and oral-motor dysfunction (Krick and Van Duyn 1984, Gisel and Patrick 1988, Rempel et al. 1988, Stallings et al. 1993, Reilly et al. 1996, Stallings et al. 1996). However, many of these studies contain only a small number of participants. In an attempt to overcome this limitation, we have recently reported the rst large-scale epidemiological study of feeding and nutritional problems in disabled children with neurological impairment, called the Oxford Feeding Study (Sullivan et al. 2000). The results of this study highlighted that feeding problems in children with neurological impairment are indeed common and severe and are a cause of much concern to parents. Patrick and Gisel (1990), on the basis of their experience with feeding neurologically impaired children, suggest that nutritional problems are caused primarily by energy decits rather than by decits of protein, vitamins or minerals. They maintain that it is because the diet given to people with disabilities is qualitatively good but decient in volume and that the malnutrition found in individuals with disabilities is rarely associated with either skin changes or hair or mucosal changes. This is, of course, quite different from the clinical picture of childhood malnutrition seen in the developing world where micronutrient deciencies (especially of vitamin A and zinc) together with protein deciency are much more prominent. The qualitative adequacy but quantitative deciency of the diets provided for children with disabilities might also be one of the reasons that such individuals can survive in a moderately malnourished state for many years. The aims of the present study were: (1) to test the hypothesis that the diet in children with disabilities is lacking in quantity rather than quality and (2) to demonstrate that the dietary decit and associated growth impairment are worse in those with more severe levels of motor decit. Method A description of the study population and the selection procedure for the children studied is provided in detail elsewhere (Sullivan et al. 2000). Briey, a target population of neurologically impaired children with oral-motor dysfunction was identified using the Oxford Register of Early Childhood Impairments (ORECI). All had been born to mothers resident in the four counties of Berkshire, Buckinghamshire, Northamptonshire and Oxfordshire in England at the time of delivery. In this cross-sectional study, 377 children (aged 4 to 14 years), born between 1984 and 1995, with motor or feeding problems recorded on the ORECI database were selected for detailed analysis. A questionnaire was sent to 377 parents of these children and replies were received from 271 (72%). The questionnaire consisted of 26 questions relating to the general health of the child, the childs nutritional status, feeding abilities, the inuence of feeding impairment on the family, and contact with health care professionals. The questionnaire was constructed in such a way that it was possible to investigate the relationship between the degree of motor disability (e.g. mobility) and oral-motor impairment (e.g. drooling and speech impairment) and difculties encountered with feeding. In the original study questionnaire (Sullivan et al. 2000), neurological impairment in participants was classied

Developmental Medicine & Child Neurology 2002, 44: 461467 461

according to type and mobility. Difculty was graded as mild (little or no difculty walking), moderate (difculty walking but does not need aids or a helper), and severe (needs aids and/or a helper or cannot walk). In this follow-up study, one hundred of the 271 families who returned the questionnaire were then randomly selected to represent a subgroup of the study population. The random selection was stratied so that equal numbers were selected from each of the four counties. After giving their consent, families of the subgroup (n=100) were visited at home by a research dietitian trained in qualitative research techniques and anthropometry. The aims of the home visits were (1) to conrm by interview answers given in the original study questionnaire, (2) to undertake a detailed dietetic analysis, and (3) to perform a comprehensive anthropometric assessment of each affected child. Approval for this project was obtained from each of the seven Clinical Research Ethics Committees covering the geographical area under study. Permission to use the ORECI database was obtained by the ORECI Management Group. After consent from the participants family doctors and informed consent from parents/caregivers had been obtained, the study questionnaire (which also asked whether or not they would give consent to be visited at home) was sent to parents/caregivers of children in order to characterize the nature and extent of the feeding disability.
DIETETIC ASSESSMENT

Health (1991) Dietary Reference Values for Food Energy and Nutrients for the United Kingdom.
ANTHROPOMETRY

Dietary information was obtained from parents in two ways. Firstly, a verbal dietary history was taken to establish the childs normal nutritional intake. This included details of foods and drinks typically eaten, quantities consumed, feeding habits and methods, information about items of food and drink taken on a more irregular basis, and a 24-hour dietary recall by a dietitian during a home visit. Secondly, all parents were asked to keep a 3-day dietary diary (Gersovitz et al.1978, Acheson et al. 1980). This diary was specically designed for the study and had simple written instructions to aid completion. Parents also received detailed verbal instructions on how to ll in the diary and they were encouraged to write down as much information as possible about the foods and drinks presented to and consumed by the child. Detailed information was requested on food brand names used, methods of cooking, and preparation and food consistency. Estimates of the amounts of food or drink (those actually ingested by the child as well as those offered in an attempt to get some idea of food loss through spillage) were recorded using handy household measures and weights (e.g. cups, uid ounces, tablespoons). Other data collected included the time in minutes taken for the child to consume each meal, snack, or drink and the extent to which the child was able to self-feed. Two weekdays and one weekend day were used in the recording period to allow for any potential differences in food consumption patterns on different days of the week. The verbal information obtained during the visit and in particular the 24-hour dietary recall was used to help validate the information in the 3-day diet diary. For those children whose parents failed to return a completed food diary, the 24-hour recall data were analyzed and used to provide information on their nutritional intakes. Dietary intake data were analyzed using proprietary software (Dietplan 5, Foresteld Software Limited). Nutrient intake was expressed as a percentage of the Department of

Detailed anthropometry was performed and included measurement of body weight (Marsden Wheelchair Weigher, Marsden Weighing Machine Group Ltd., London), upper arm and lower leg length (Harpenden Anthropometer, Child Growth Foundation, London), mid-upper arm circumference (MUAC; Harpenden Anthropometric tape), triceps and subscapular skinfold thickness (skinfold callipers, Holtain Ltd., Crymmych, UK), and occipito-frontal circumference (Lasso Headtape Measure, Child Growth Foundation, UK). To reduce the potential for measurement error, all measurements were performed by the same trained operator and each measurement was repeated three times consecutively and the mean value calculated. Accurate length and height measurements are difcult to obtain in children with disabilities because of limb contractures and scoliosis. For this reason, we chose to record upper arm length and lower leg length (Spender et al. 1989) as measures of linear growth. In order to identify possible differences in limb length and body symmetry in those children with a right or left-sided hemiplegia, measurements were taken from both the right and left sides of the body. However, to make the Table readable, we only present results for left-sided measurements (correlation between left and right-sided measurements was very high: the lowest Pearsons correlation coefcient was 0.92 for left and right upper arm). In order to facilitate internal and external comparisons adjusting for the known confounding effects of age and sex, measurements were expressed as standardized z scores of available ageand sex-specic reference population standards (a negative z score indicates below average). For weight and occipito-frontal circumference, measurements were standardized to the 1990 British Growth reference centiles (Freeman et al. 1995), upper arm and lower leg length measurements were standardized to American reference standards for cerebral palsy as none are available for British children (Spender et al. 1989), and mid upper arm circumference and skinfold measurements were standardized to American data (Frisancho 1988).
STATISTICAL ANALYSIS

Descriptive and statistical analyses were performed using SPSS (version 10.0) and STATA (version 6.0). The continuous variables of age, weight, and standardized z scores were investigated for departure from normality both informally, by assessing the shape when looking at a histogram, and formally, using the Shapiro-Wilk test for normality. The majority of children were in the lowest centile for their measurements and, therefore, the data were, by nature, skewed. Having examined the histograms and tests of normality, we decided to present descriptive statistics appropriate for skewed data, and perform non-parametric tests throughout to investigate differences between the disability groups with respect to demographic, ponderal (weight, MUAC, triceps and subscapular skinfold thickness), and skeletal measurements. For the continuous variables of age and weight, summary statistics including the median and the range are given (Table II). In addition, for the standardized z scores, we present the proportion of children below a dened range expressed as a percentage (Table II). The lower end of the dened range was a z score of 1.96 (i.e. 2.5% of values were expected to lie

462

Developmental Medicine & Child Neurology 2002, 44: 461467

below this level) and the denominator for this percentage was as the number of valid measurements. We acknowledge that 1.96 is an arbitrary cut-off (representing approximately 2SDs below mean) and that some children did indeed register measurements above a z score of +1.96; however, the focus of this study is the lower end of the weight spectrum and hence a one-sided aspect is presented here. Comparison of age and standardized z scores across all three disability groups was performed using the Kruskal-Wallis (equality of populations rank) test. Post-hoc comparisons contrasting mild and moderate disability combined versus severe disability were also performed using the Kruskal-Wallis test, equivalent to performing a Mann-Whitney U test. For the binary variable, sex, the distribution of males and females across the three disability groups was investigated using the 2 test. Results The hundred children with disabilities visited in their homes were evenly spread across the four counties of Berkshire, Buckinghamshire, Northamptonshire, and Oxfordshire. In the vast majority of cases, the caregiver was the childs own mother (in four cases it was a foster mother). The mean age of the children was 9 years (SD 2 years 5 months), ranging from 4 years 6 months to 13 years 7 months (40 females 60 males). There was no statistically signicant difference between the disability groups with respect to the distributions of age and sex, that is, the groups were well balanced for these characteristics. Forty-eight children were unable to walk and 21 needed both aids and a helper to be mobile; these 69 children were classied for the purposes of analysis as the severely disabled group. Nineteen children had moderate difculty with walking, but no aids nor helpers were required for them to be mobile (moderately disabled group). The remaining 12 children had either little or no difficulty with walking (mildly disabled group). Cerebral palsy (CP) was the reason for disability in 90 children, of whom 64 (71%) had spastic quadriplegia. The remainder of the diagnostic categories were as follows: global developmental delay (n=3), Marfan syndrome (n=1), intractable epilepsy (n=2), agenesis of the corpus callosum (n=2), methyl malonic aciduria (n=1), and congenital rubella (n=1). In terms of responses to the study questionnaire, when compared with children in the total study population (n=271), those children in the home visit study (n=100) were not substantially different in terms of age, sex, or geographical locality within the region. Nor were they different in the important areas of motor impairment, oral-motor impairment, feeding problems, and nutrition;

therefore, we consider them a representative sample of the total population of children with CP with or without motor dysfunction. Table I shows the relationship between ability to self-feed and disability and indicates that the majority of children with a severe degree of motor impairment were unable to selffeed and required help. All of the children with a mild to moderate degree of motor impairment were able to be fed by mouth and none always needed help with feeding.
DIETETIC ASSESSMENT

Fifty-three caregivers returned dietary diaries. The verbal diet history information was used to provide information on nutrient intakes in 41 participants. Six children, all with severe disabilities, were given proprietary enteral feeds via their gastrostomy tubes and in these cases nutritional intake was calculated from information provided by the product manufacturers. Figure 1 shows nutrient intakes with respect to the percentage of children achieving over 100% of estimated average requirement (EAR). Nutrient intake analysis revealed that energy intakes were generally low with only 20% of children having energy intakes above the EAR for their age and 59% of the group with severe disabilities consuming below 80% EAR compared with 16% of the group with moderate disabilities, though this difference was not statistically significant (Kruskal-Wallis test comparing mild plus moderate vs severe disability, p=0.09). Ninety-six out of 100 children, however, achieved a protein intake above the EAR and this was consistent across the three groups. Mean intake of the micronutrients: sodium, potassium, magnesium, phosphorus, copper, iodine, B1, B2, B6, and B12, met reference nutrient intake (RNI) for all children and for each disability group, but the ranges of intakes were wide. It is noteworthy that iron, for instance, fell below the normal range for nearly half the participants. Interestingly, in 22 out of 33 of the children with severe disabilities, milk or milk-based drinks provided the highest proportion of energy in their diets out of all the items they consumed. Milk is a poor source of iron and may explain this nding together with the observation that the majority of these children met their RNIs for calcium, phosphorus, thiamin, riboavin, B6, and B12 more than adequately. Half of the children with severe disabilities failed to meet at least 81% of the RNI for potassium, iron, copper, magnesium, and zinc. Selenium, vitamin A, niacin, and folate deciencies were also seen in moderate and mild disability groups. Retinol (vitamin A) intakes were relatively low across all the groups with only 26% of children meeting the RNI.

Table I: Feeding ability with respect to degree of motor disability, n (%)

Severe disability (n=69) Always needs help Some difculty, needs help Can feed but slow and messy, help given Tube fed/not fed by mouth Some difculty, no help Self feeds 33 (48) 15 (22) 12 (17) 6 (9) 2 (3) 1 (1) Mild or moderate disability (n=31) 0 (0) 23 (74) 2 (6.5) 0 (0) 4 (13) 2 (6.5)

Nutritional Intake and Growth in Disability P B Sullivan et al.

463

Table II: Summary statistics for continuous variables (z scores are standardized for age and sex): median, proportion (%) of children more than 2SDs below mean (for z scores only) and range
Index All children n=100 Mild n=12 9.63 (6.2 to 12.8) 29.45 (12.7 to 43.5) 0.31 (8%) (4.8 to 1.6) 0.45 (25%) (3.3 to 0.8) 0.55 (0%) (1.5 to 2.2) 0.000 (0%) (1.4 to 1.8) 0.39 (0%) (1.4 to 1.9) 0.25 (0%) (1.5 to 2) 0.18 (0%) (0.8 to 1.9) Disability group Moderate n=19 9.33 (6.4 to 13.3) 25.90 (17.8 to 66.4) 0.62 (16%) (2.1 to 2) 1.76 (33%) (3 to 0.8) 0.90 (0%) (0.8 to 2.9) 0.27 (5%) (2.1 to 3.6) 0.70 (0%) (1.5 to 2.1) 0.25 (0%) (1.2 to 2.3) 0.063 (0%) (0.8 to 1.7) Severe n=69 8.92 (4.5 to 13.6) 20.40 (11.1 to 51.2) 1.87f (49%) (10.2 to 1.7) 2.26h (56%) (7.3 to 1.9) 0.35f (9%) (3.5 to 4.7) 2.08f (52%) (5.9 to 2.8) 0.74 (13%) (3 to 2.4) 0.89g (2%) (2.1 to 2.9) 0.40h (0%) (1.2 to 3.9)

Age (y) Weight (kg) Weight, z scorea Occipito-frontal measurement, z scorea Upper arm length, z scoreb Lower leg length, z scoreb Mid upper arm circumference, z scorec Triceps skinfold, z scorec Subscapular skinfold, z scorec

9.04 (4.5 to 13.6) 22.60 (11.1 to 66.4) 1.43d (38%) (10.2 to 2) 1.92 (48%) (7.3 to 1.9) 0.000d (6%) (3.5 to 4.7) 1.27d (36%) (5.9 to 3.6) 0.69 (9%) (3.5 to 2.4) 0.63e (1%) (2.1 to 2.9) 0.29 (0%) (1.2 to 3.9)

aReference population,1990 British Growth reference centiles (Freeman et al.

1995); bReference population, Spender et al. 1989;

cReference population, Frisancho 1988; dp 0.001; ep 0.05; results of Kruskal-Wallis test across all 3 disability groups; fp 0.001; gp 0.01; hp<0.05; post-hoc contrasts revealed signicant differences when severe group was compared with combined mild and moderate groups.

100 90 96 85 77 70 60 50 40 30 20 20 10 0 26 54 46 52 44 43 71 62 96 86 89

Percentage above 100% EAR or RNI

80

te la Fo vin a bo Ri e in m ia Th ol tin Re um ni le Se

nc Zi us or ph os Ph m siu ne ag M C

in m ta Vi

Figure 1: Dietary intakes of children visited at home. Estimated average requirement (EAR) Reference nutrient intake (RNI)
464 Developmental Medicine & Child Neurology 2002, 44: 461467

n Iro um lci Ca m iu ss ta Po um di So n ei ot Pr gy er En

ANTHROPOMETRY

Table II shows summary statistics for the measurements recorded for the whole group of participants and with respect to level of motor disability. Figures 2 to 5 examine the relationship between level of motor dysfunction and standardized body weight, lower leg length, upper arm length, and occipito-frontal measurement (head circumference) respectively, using dotplots. Dotplots present all the observations, in this case z scores by disability group, with the added features of a horizontal line representing the mean value for each reference population (i.e. z score=0), and short horizontal line indicating the median value for each disability group. Over three-quarters of children in this study group were below mean weight for their age and sex when compared

with the standards for children without disabilities and this decit was most marked in those with severe motor disability where their median value approximates to 2SDs below the reference mean value (Fig. 2). Linear growth, as measured by limb length, was similarly impaired and it is noteworthy that the compromise in linear growth was more marked in the lower limbs and almost exclusively in the severe disability group. For lower leg length, the mild and moderate disability groups had typically normal measurements, while in the severe disability group, half were more than 2SDs below the reference mean value (Fig. 3). In the upper limb, around three-quarters of the mild and moderate disability groups were above the reference population mean (Fig. 4), whereas in the severe disability group, the

3 2 1 0 1 2 3 4 5 6 7 8 9 10 Mild Moderate Severe

5 4

Upper arm length z score

3 2 1 0 1 2 3 4 Mild Moderate Severe

Weight z score

Disability Group Figure 2: Dotplot of bodyweight z score by disability group. Line indicates median value for each disability group.

Disability Group Figure 4: Dotplot of upper arm length z score by disability group. Line indicates median value for each disability group.

4 3

3 2

Head circumference z score

Lower leg length z score

2 1 0 1 2 3 4 5 6 Mild Moderate Severe

1 0 1 2 3 4 5 6 7 Mild Moderate Severe

Disability Group Figure 3: Dotplot of lower leg length z score by disability group. Line indicates median value for each disability group.

Disability Group Figure 5: Dotplot of head circumference z score by disability group. Line indicates median value for each disability group.

Nutritional Intake and Growth in Disability P B Sullivan et al.

465

median z score was 0.35 when compared with the Spender (1989) reference population. It should be remembered here that this reference population for linear growth is a CP population (Spender 1989) and thus this growth distinction between mobile and immobile children with CP is all the more marked. Head growth (Fig. 5) was also reduced in children with neurological impairment and worsened with severity of motor disability such that the median z scores for those with mild, moderate, or severe motor disability were 0.5, 1.8, and 2.3, respectively. Another way of looking at the measurement of head circumference is that, overall, nearly half of all children (n=48) were below our dened threshold of 2SDs below the mean, with a trend according to severity of disability: 1 in 4 for the mild disability group, 1 in 3 for the moderate disability group, and 1 in 2 for those with severe motor impairment. Ponderal measurements, that is, weight, MUAC, triceps and subscapular skinfold thickness, were most markedly atypical in proportion to the degree of motor disability; weight for age, for instance, was below the reference range in 8%, 16%, and 49% of participants for the mild, moderate, and severe disability groups, respectively (Table II). That is, for the most severely disabled group, we would expect 2.5% of observations to lie below our dened range, however, 49% of observations were below this level. Moreover, when post-hoc comparisons were made for z scores across disability groups for the ponderal measurements, the difference between mild and moderate combined versus severe disability groups was statistically signicant for weight (p<0.001), triceps (p<0.01), and subscapular skinfold thickness (p<0.05; Table II). For measurements of linear growth, i.e. upper arm and lower leg length measurements, the highly statistically signicant differences between z scores in the three disability groups were almost entirely due to the low scores of the severely disabled group. Post-hoc comparisons of z scores showed that the difference between mild and moderate combined versus severe disability groups was highly statistically signicant for both lower leg length (p<0.001) and upper arm length (p<0.001). Discussion
ENERGY INTAKE

to growth impairment in children with disabilities (Stevenson et al.1995), it would appear that for many children a suboptimal energy intake does contribute to poor growth and low weight for age.
MICRONUTRIENT INTAKE

We have demonstrated that feeding and nutritional problems are common in children with disabilities with neurological impairment (Sullivan et al. 2000). Moreover, such problems are probably under recognized and are a source of distress for both children and parents; for example, 93 out of 240 (39%) parents considered their child to be underweight and, in a similar nding to that of Reilly and Skuse (1992), 43% (113 out of 264) described feeding their child with a disability as stressful and not enjoyable. Results of the present study conrm the view of Patrick and Gisel (1990) and show that the majority of children (80 of 100 children) consumed a diet that was decient in energy when compared with that usually required by non-disabled peers. Energy intake is, to some extent, linked to mobility and activity level. It might, therefore, be expected that children with disabilities would have lower energy requirements than children without disabilities. Moreover, the precise nutritional requirements for children with disabilities are unknown and the reference standards used to gauge intake may be inappropriate for this group of children. Nevertheless, although it is likely that that there are non-nutritional factors contributing

Micronutrient deciencies, including those of calcium, iron, zinc, vitamin C, vitamin A, riboavin, and thiamine, have been reported in children with developmental disability and may be under recognized (Garty et al.1989, Patrick and Gisel 1990). For instance, children with psychomotor disability may be at a higher risk of subclinical vitamin C deciency or overt scurvy (Front et al.1978, Garty et al.1989). Mean micronutrient intakes of the present study group appeared to be adequate (with the exception of Retinol), when compared with the daily-recommended values for children without disabilities. However, poor dietary intake and nutritional status were often observed on an individual level and especially in children with severe disabilities. The choice of foods obviously inuences the balance of nutrients present in the diet. Milk and milk based drinks provided the highest proportion of energy for most of the children with severe disabilities and thus the intake of micronutrients was biased toward those abundant in dairy products. Participants diets were most decient in nutrients found in foods excluded from the diet because they were difcult for the child to eat. Meat (a rich source of iron and zinc), fruits, and vegetables (sources of vitamin C) were reported by parents as items often avoided because of feeding problems. Children with disabilities may be at greater risk of iron deciency as their diet is often limited and especially so if the diet is based largely on cows milk. Results from the present study show that iron intake is suboptimal in many children with disabilities, with only 46% achieving 100% of their RNI. For those participants who had an adequate iron intake, this probably reects their frequent consumption of fortied breakfast cereals. There is now sufcient evidence that iron deciency is detrimental to neurodevelopmental progress and that if left untreated deciencies will persist; an irondecient state at the time of rapid brain growth does appear to impair brain function and such effects may remain (Palti et al.1985). Conversely, treatment of iron deciency can produce a rapid improvement in developmental scores in infants. Idjradinata and Pollitt (1993) showed that although iron-decient anaemic infants perform worse in tests of motor and mental development than do iron-sufcient infants of a comparable age, such developmental delay could be reversed by treatment with ferrous sulphate.
GROWTH AND ANTHROPOMETRY

Results from this study conrm the signicant adverse effect of CP on linear growth and also reveal the differential growth effect with the lower limb being more profoundly affected than the upper limb. This presumably relates to the importance of mobility and weight bearing in stimulating leg growth as the effect was signicantly more marked in the group with severe motor-impairment (Fig. 3). Weight and mid upper arm circumference measurements are easy to perform and in this study gave a good indication of nutritional status. Stunting and wasting were prevalent, especially in the children with severe disabilities, but were also observed at an individual level in those with mild and moderate mobility decits.

466

Developmental Medicine & Child Neurology 2002, 44: 461467

Effect of severity of motor impairment on growth and nutritional intake Although we did not use a standard scale of mobility but rather based our classication on the answers given in the questionnaire in our original study (Sullivan 2000), we were nevertheless able to identify clearly those children with the most severe degree of motor disability. In a recent prospective follow-up of 321 children born with bilateral CP born in London between 1989 and 1992 (Lin et al. 1999), only 38% attained unsupported walking. Using the classication in our study, the gure was very similar with only 31% being able to walk unsupported. Conclusion This study has conrmed the signicant impact of neurological impairment in children on body growth and nutritional status. Moreover, it has demonstrated that these decits are worse in those with a greater degree of motor impairment. The major nutritional decit was in energy intake, whilst micronutrient intakes were less markedly impaired and protein intake was normal in this group of children. Many children with neurological impairment, regardless of the degree of disability, would benet from individual nutritional assessment and management as part of their overall care. Dietitians have a very important role in ensuring that this is achieved and should be actively involved in the process of nutritional assessment. Further research is required to determine the appropriate energy intake required to maintain energy balance in children with neurological impairment.
Accepted for publication 10th January 2002. Acknowledgements The authors are grateful to the managing committee of the Oxford Register of Early Childhood Impairment (ORECI) for permission to use the database. This study was made possible by a grant from the Anglia and Oxford Research and Development Committee. References American Dietetic Association. (1992) Position of the American Dietetic Association: nutrition in comprehensive program planning for persons with developmental disabilities. Journal of the American Dietetic Association 92: 6135. Acheson KJ, Campbell IT, Edholm OG, Miller DS, Stock MJ. (1980) The measurement of food and energy intake in man an evaluation of some techniques. American Journal of Clinical Nutrition 33: 114754. Department of Health. (1991) Dietary Reference Values for Food, Energy and Nutrients for the United Kingdom. Report on Health and Social Subjects No. 41. London: Her Majestys Stationery Ofce. Freeman JV , Cole TJ, Chinn S, Jones PRM, White EM, Preece MA. (1995) Cross-sectional stature and weight reference curves for the UK, 1990. Archives of Disease in Childhood 73: 1724.

Frisancho AR. (1988) Anthropometric Standards for the Assessment of Growth and Nutritional Status. Ann Arbor, MI: Health Products. Front D, Hardoff R, Levy J, Benderly A. (1978) Bone scintigraphy in scurvy. Journal of Nuclear Medicine 19: 9167. Garty BZ, Danon YL, Grunebaum M, Nitzan M. (1989) Scurvy in children with severe psychomotor retardation. International Pediatrics 4: 27982. Gersovitz M, Madden JP , Smiciklas-Wright H. (1978) Validity of the 24-hr. dietary recall and seven-day record for group comparisons. Journal of the American Dietetic Association 73: 4855. Gisel EG, Patrick J. (1988) Identication of children with CP unable to maintain a normal nutritional state. Lancet 1: 2836. Idjradinata P , Pollitt E. (1993) Reversal of developmental delays in iron-decient anaemic infants treated with iron. Lancet 341: 14. Krick J, Van Duyn MA. (1984) The relationship between oral-motor involvement and growth: a pilot study in a pediatric population with cerebral palsy. Journal of the American Dietetic Association 84: 5559. Lin J-P , Scrutton D, Baird G. (1999) The changing natural history for walking in bilateral cerebral palsy. Developmental Medicine & Child Neurology 41 (Suppl. 82): 17 (Abstract). Palti H, Meijer A, Adler B. (1985) Learning achievement and behavior at school of anemic and non-anemic infants. Early Human Development 10: 21723. Patrick J, Gisel E. (1990) Nutrition for the feeding impaired child. Journal of Neurology and Rehabilitation 4: 1159. Reilly JJ, Hassan TM, Braekken A, Jolly J, Day RE. (1996) Growth retardation and undernutrition in children with spastic cerebral palsy. Journal of Human Nutrition and Dietetics 9: 42935. Reilly S, Skuse D. (1992) Characteristics and management of feeding problems of young children with cerebral palsy. Developmental Medicine & Child Neurology 34: 37988. Rempel GR, Colwell SO, Nelson RP . (1988) Growth in children with cerebral palsy fed via gastrostomy. Pediatrics 82: 85762. Spender QW , Cronk CE, Charney EB, Stallings VA. (1989) Assessment of linear growth of children with cerebral palsy: use of alternative measures to height or length Developmental Medicine & Child Neurology 31: 20614. Stallings VA, Charney EB, Davies JC, Cronk CE. (1993) Nutritionrelated growth failure of children with quadriplegic cerebral palsy. Developmental Medicine & Child Neurology 35: 12638. Stallings VA, Cronk CE, Zemel BS, Charney EB. (1995) Body composition in children with spastic quadriplegic cerebral palsy. Journal of Pediatrics 126: 8339. Stallings VA, Zemel BS, Davies JC, Cronk CE, Charney EB. (1996) Energy expenditure of children and adolescents with severe disabilities: a cerebral palsy model. American Journal of Clinical Nutrition 64: 62734. Stevenson RD, Roberts CD, Vogtle L. (1995) The effects of nonnutritional factors on growth in cerebral palsy. Developmental Medicine & Child Neurology 37: 12430. Sullivan PB, Lambert B, Ford-Adams M, Grifths P , Johnson A. (2000) Prevalence and severity of feeding and nutritional problems in children with neurological impairment: Oxford Feeding Study. Developmental Medicine & Child Neurology 42: 67480. Thommessen M, Kase BF, Riis G, Heiberg A. (1991) The impact of feeding problems on growth and energy intake in children with cerebral palsy. European Journal of Clinical Nutrition 45: 47987.

Nutritional Intake and Growth in Disability P B Sullivan et al.

467