J. Paediatr.

Child Health (1990) 26, 335-338

Nutritional assessment of children with

juvenile chronic arthritis
A. L. MORTENSEN, J. R. ALLEN* and R. C. ALLEN
The Children's Hospital, Camperdown, NSW 2050, Australia

Abstract Nutritional assessment was performed in 38 children with juvenile chronic arthritis (polyarticular [15]; pauci-
articular [13]; systemic [lo]).This included anthropometric measurements (height, weight, mid-upper arm circumference and
four skinfolds) and a 7 day weighed food record. The children were classified according to current type of arthritis which in
some cases was not the same as the initial onset type. Mean height Z scores were significantly below the Zscore population
mean of 0 in the systemic (P = 0.02) and polyarticular (P = 0.009) groups. Mean weight Zscores were also below the mean
in the polyarticular group (P= 0.001) but the systemic group did not reach significance. Mean energy intakes were
significantly below the recommended dietary intake (RDI) in the systemic (P = 0.01) and polyarticular (P = 0.001) groups.
Mean intakes of calcium and zinc were below the RDI of 100% in the polyarticular group (P = 0.0001). Thirteen (34%) of
children were taking some form of self-prescribed vitamin and/or mineral supplement. Vitamin C was the most commonly
used supplement, despite dietary intakes greatly exceeding the RDI. Nutritional assessment is essential in the management
of patients with JCA, in particular those with systemic or polyarticular disease.

Key words: growth; juvenile chronic arthritis; nutritional assessment,

Juvenile chronic arthritis (JCA) affects approximately 1 in 2000
children under 16 years of age.' Classification criteria are based
predominantly on clinical parameters of joint count and extra-
articular features occurring during the initial 6 months of disease
and include pauci-articular (four or less joints), polyarticular
(five or more joints) or systemic onset (arthritis with fever and
rash).2 Although the majority of children have pauci-articular
onset JCA, which has a generally good prognosis, patients with
systemic or polyarticular onset and those who progress from a
pauci-articular to polyarticular course may have significant
disease sequelae.
Poor somatic growth has long been a recognized feature of
JCA, the 'general arrest of development' being specifically noted
by Still in his initial report of 1897.3 The majority of reported
studies in JCA relate poor growth to adverse effects of
corticosteroids such that possible primary disease related
aspects have been inadequately studied.* No consistent growth
hormone abnormality has been found, irrespective of corti-
costeroid therapy, despite low levels of insulin-like growth
factor (IGF)/somatomedin being reported5 Adequate nutrition
is essential for normal growth yet few reported studies have
addressed this aspect in JCA particularly in the young child, at
which time disease effects on growth may be critical.6.' IGF-1 is
low in malnourished children, without associated endocrine
abnormalities, yet normalize with nutritional rehabilitation and
Correspondence: Or R. C. Allen, The Children's Hospital, Camperdown.
NSW 2050, Australia.
'Present address: J. R. Allen, James Fairiax Institute of Paediatric
Clinical Nutrition, The Children's Hospital, Camperdown, NSW 2050,
Australia.
A. L. Mortensen, BSc,DipNutr&Diet, Research Paediatric Dietitian; J. R.
Allen, MSc, DipNutr&Diet,Research Paediatric Dietitian; R. C. Allen, MB,
BS. FRACP, Paediatric Rheumatologist.
Accepted for publication 22 September 1990.
this supports the need to evaluate nutritional status and growth
in children with JCA.' The aim of this study was to assess the
nutritional status and dietary adequacy in children with JCA.
METHODS
Patients
Thirty-eight patients with a diagnosis of JCA based on the
EULAR2 classification were studied: 13 pauci-articular, 15
polyarticular and 10 systemic onset. The patients were recruited
from the paediatric rheumatology clinic of The Children's
Hospital, Camperdown, as well as referred from other
rheumatology services. All patients from the clinic were enrolled,
pending parental consent. irrespective of concern relating to
nutrition and growth. Four pauci-articular patients had pro-
gressed to a polyarticular course and were included with the
polyarticular onset subgroup. Of the systemic onset patients,
although some followed a pauci-articular course, because of
the distinctive nature of the systemic-onset subgroup they have
been analysed independently. One child with pauci-articular
onset disease, of short duration, was withdrawn from the study
because his growth and nutritional status were probably com-
promised by extreme social circumstances, including an
extensive period in a refugee camp before resettlement in
Australia.
All the children had active disease at the time of entry in the
study based on the presence of clinical synovitis. periods of
morning stiffness and functional impairment relative to the
involved site. Acute phase reactants, including erythrocyte
sedimentation rate and C-reactive protein, were elevated in
most patients, particularly polyarticular and systemic onset, but
were not sequentially followed for purposes of this study. All
patients required non-steroidal anti-inflammatory drug (NSAID)
336 A. L. Mortensen eta/.

therapy, of whom 14 also received slow-acting anti-rheumatic Table 1 Characteristics of groups by disease type
drugs (SAARD). two cytotoxics and 12 corticosteroids. The
anthropometric measurements were height, weight and skinfold Pauci Systemic Polyarticular
thicknesses at four sites. Height was attained with a stadiometer n=13 n = 10 n = 15
(to the nearest 0.1 cm) following Cameron’s technique9 Age (years) 7.4f4.4 6.8f 4.4 9.1i4.3
An electronic platform beam scale (accurate to the nearest Sex (M:F) 3:lO 5:5 5:lO
100 g) was used to weigh all the children in minimal clothing. An Disease duration (years) +
2.5 2.4 *
2.7 2.9 3.0f 2.2
average of three skinfold measurements was taken with Holtain
skinfold calipers (to the nearest 0.1 mm) at the tricep, bicep, Values are expressed as mean and s.d. except for sex.
subscapular and supra-iliac sites. Mid-upper arm circumference
was measured with a firm tape to the nearest 0.1 cm. All
Table 2 Anthropometric variables of groups by disease type
measurements were taken on the left hand side, eight children
refused or were too restless to have skinfold measurements Pauci Systemic Polyarticular
taken, and two had only the triceps and mid-upper arm circum-
ference measurement. 2 height score -0.16f1.0 -1.1 fl.l* -0.75+0.96**
The height and weight for each child was standardized for Z weight score +0.18f 1.19-0.39k 1.2 -1.07 f l.O***t
age and sex against normalized international reference values Per cent ideal weight for
assembled by the US National Center for Health Statistics.” The height age 105f10 109f16 94*11
Mid upper arm muscle area,
standardized score (Z score) was calculated by the equation:
<5th percentile 1/11 216 4/13
Z = y-x/s.d. of x, where y was the child’s height or weight, x
Mid upper arm fat area,
was the median height or weight of a child of the same age and <5th percentile 311 1 216 6/13
sex and s.d. of x was 1 s.d. above or below the median reference
value. P<0.05. ** P<O.Ol.* * * P<O.OOl (Student’s t-test: difference
Percentage body fat was calculated using either Durnin and in Zscores from the population mean).
+ P<O.OOl (Mann-Whitney U test: difference between the pauci- and
Rahamin’s” or Brook’s’2 equation according to the age of the
child. The mean triceps skinfold measurement, upper arm muscle polyarticular groups).
Values are expressed as mean and s.d. except for mid upper arm
and fat area were compared with Frisancho’s fifth percentile
muscle area and mid upper arm fat area.
values for children of the same age and sex.13

Dietary analyses
Parents were instructed to keep a 7 day record of their child’s
food intake. Older children kept their own record. All foods and
drinks consumed during this time were to be weighed with a
portable set of battery operated scales (accurate to nearest 5 9).
Recipes were written on a separate sheet with all the weights of
the raw ingredients and the total cooked weight to allow for the
calculation of water loss during cooking. Any food items which
were not weighed were recorded in descriptive metric measure-
ments. All information was recorded in a food record book,
which was checked during and at the end of the week by a
dietitian (A.M. or J.A.). Medications, vitamin and mineral supple-
ments used were recorded and checked. The food records
were analysed with a computerized program Diet 1 (Xyris
Software, Queensland) of the Australian food composition tables
Nuttab 1988. The micronutrient intakes were compared with the
Australian recommended dietary intakes (RD1).14 Energy and
protein intakes were compared with FAO/WHO requirements
based on each child’s actual weight or ideal weight for height
age.’5
Statistics
The Minitab statistical computer package (Siromath Pty. Ltd,
Sydney) was used for all analysis and the rejection level was set
at 0.05. The significance of intergroup comparisons was cal-
culated with ANOVA and Mann-Whitney U test. Student’s t-test
was used to check for any significant difference in Z scores
from the population mean of 0, and nutrient intake from
recommended dietary intake (RDI) of 1001.
RESULTS
The characteristics of the groups based on disease type are
shown in Table 1. There was no significant difference in age,
male to female ratio and disease duration between the groups.
Anthropometric variables are shown in Table 2. Mean height Z
scores were significantly below the Zscore population mean of
0 in the systemic ( P = 0.02) and polyarticular ( P = 0.009) groups.
Mean weight Z scores were also below the mean in the poly-
articular group (P = 0.001) and in the systemic group, although
this did not reach significance. The mean weight Zscore in the
polyarticular group was significantly less than the pauci group
( P = 0,008).There was no significant difference in the mean
weight for height index between the groups and they were all
within the normal range (90-1 10%). One-third of the subjects in
the systemic and polyarticular groups had mid-upper arm muscle
area less than the 5th percentile. One-third of the subjects in the
systemic and one-half of the polyarticular group had mid-upper
arm fat area less than the 5th per~enti1e.l~ Corticosteroid use
was significantly associated with low Z height scores
(Chi-squared test, P = 0.005).
Comparison of nutrient intake by disease type is shown in Fig.
1. Mean energy intakes were significantly below the RDI in the
systemic ( P = 0.01) and polyarticular ( P = 0.001) groups. The
mean protein intakes considerably exceeded the RDI in all
groups. The percentage of energy contributed from carbohydrate
(4941%). fat (34-36%) and protein (14-17%) was similar in all
groups.
The values given for vitamins and minerals exclude supple-
ments. Mean intakes of calcium and zinc were below the RDI of
100% in the polyarticular group ( P = 0.001). The mean intakes
for iron, thiamin, niacin equivalents, riboflavin, Vitamins C and A
were all above the RDI in each group. As thiamin, ,(acin and
JCA and nutritional assessment 337

400 lDIIIlD Pauci

0Systemic
300
rn Poly
-
n
I 200
I
c
-
.-m
!?
CI
v)

8 100

Energy Prot Ca Fe Zinc Vit C Vit 81 Vit 82 Niacin Vit A

Fig. 1 Comparison of mean nutrient intake, expressed as percentage of Australian RDI for each disease subgroup. Prot = protein; Ca = calcium;
Fe = iron; Vit 61 = Vitamin 61 (thiamin):Vit 62 = Vitamin 6 2 (riboflavin); Vit A = Vitamin A (retinol equivalents).

riboflavin are involved in energy metabolism, the requirement is
usually related to energy intake. The nutrient density of thiamin,
riboflavin and niacin equivalents (expressed as intakeM.2 MJ
(1000 kcal)) exceeded the RDI in all groups (data not shown).
Although not included in the nutrient analysis, 13 (34%)
children were taking some form of self-prescribed supplement.
Ten children were taking the supplements daily, while three
were taking them irregularly. Vitamin C was the most frequently
used supplement, being taken by six children on a daily basis.
Vitamin C doses ranged from 300 up to 1500 mg in excess of the
RDI of 30 mg. Multivitamin preparations were taken by seven
children including iron-fortified supplements in four. Homeo-
pathic remedies or cod-liver oil were each taken by two children
daily.
DISCUSSION
Adequate nutrition is essential for normal growth, yet few studies
have addressed this in relation to the well-recognized feature of
poor somatic growth in children with JCA. In this study, we set
out to assess the nutritional status and dietary adequacy of a
group of children with JCA and thereby determine whether
nutritional factors may in part explain the poor growth and
anthropometric parameters observed.
Retardation in both height and weight was evident in systemic
and polyarticular disease. Wasting of muscle and fat stores
being demonstrated in one-third of these groups. Similar findings
were reported by Johansson with subnormal height, triceps
skinfolds, and arm muscle circumference in teenage girls with
systemic or polyarticular disease.I6 Lovell et a/. also reported
growth failure and, based on the weight for height index,
moderate to severe wasting in half their group with JCA.’ We
did not find the weight for height index useful in describing the
wasting demonstrated in that only three fell into the range of
severe to moderate reduction whereas using the standardized Z
score, eight were beyond 2 s.d. for height and/or weight. That
suggests that this index does not accurately reflect the nutritional
status of our groups, particularly where proportional reduction
in somatic growth is present. A further reason for this is the
difficulty in measuring height accurately due to flexion deformity.
Many studies describe the adverse effects of corticosteroid
use on linear growth.” Corticosteroid use was significantly
associated with low Z height scores in our group (Chi-squared
test, P = 0.005). However, low Zweight scores and both muscle
and fat wasting was also evident in the systemic and polyarticular
groups in which corticosteroid use was common (70 and 60%
respectively). This suggests other factors are involved, given the
appetite stimulant and weight gaining effects of corticosteroids.
Disease activity may be one factor in that when inflammatory
activity is high, particularly a flare of systemic disease, appetite
is generally poor and thus nutrient intake is compromised.
We are aware of only one other study which has evaluated
the nutritional adequacy of the diets of children with JCA by
Portinsson et a/. who did not find any evidence of low nutrient
intakes in their study of teenage girls with JCA. compared with
Swedish RDI and controls, although zinc intake was not
reported6 In comparison, our study has shown significantly low
energy intakes in the systemic and polyarticular groups and
also low zinc and calcium intakes in the polyarticular group.
These differences may be due to three main reasons. First, our
dietary data were collected from 7 day weighed food records, in
comparison with a 4 day record by household measures. This
338
method was chosen as it gives the most accurate picture of an
individual’s average nutrient intake.ls Although it could be
argued that the arduous task of weighing may alter an indi-
vidual’s usual intake, this was probably not the case in our
population as it was emphasized to the parents the importance
of not altering their children’s diet. There was no reduction in
energy intake over the 7 days, which indicates no deterioration
in record keeping. Second, our patient population included
more polyarticular and systemic patients (76% in comparison
with 50%), in whom lower dietary intakes were observed and
therefore the two studies are not entirely comparable. Finally,
our population covered a wide age range of both sexes (18
months-1 7 years) as compared with only females aged 11-16
years. We had insufficient numbers to group by age as well as
disease onset type and. therefore, cannot exclude the possibility
that particular age groups are closer to the RDI: however our
overall impression was that poor intake was not an infrequent
finding in the teenager.
One reason for low calcium intakes was a high consumption
of fruit juice particularly in the toddler age-group, which dis-
places cow’s milk from the diet-the major source of calcium in
this age group. Also, five children had been given a trial of
milk-free or low dairy product diets on the advice of natural therapists
and ‘self-help’ books, without any appropriate cow’s milk sub-
stitutes being used. Poor zinc intakes were associated with
poor energy intakes and in many cases, were associated also
with low calcium intakes as milk and dairy products provide
approximately 20% of zinc in Australian children’s diets.I9 The
current RDI for zinc are under review by the NHMRCig due to
recent evidence of increased rate of absorption of this nutrient
in humans, thus our groups intakes may not be as low in terms
of the RDI as reported. It is interesting to note that the three
nutrients that are low are important factors for normal growth.20
Gorticosteroids increase the urinary excretion of zinc and
calciumz1 and thus requirements would theoretically be higher
in children taking this therapy. The polyarticular group appears
to be particularly at increased risk of calcium and zinc deficiency
since their intake was significantly low and corticosteroid use is
common.
Although self-prescribed vitamin supplementation was
common, no calcium and zinc supplements were being taken
by children with low dietary intakes. Vitamin C was the most
commonly used supplement, doses ranging from 500 up to 1500
mg, despite dietary intakes greatly exceeding the RDI of 30 mg.
Toxic side effects (renal stones, rebound scurvy), have been
reported at intakes greater than 500 mg dailyzz which indicates
the necessity to assess supplement use and dosage, considering
the potential contraindications. The multivitamins taken did not
contain calcium or zinc and supplied vitamins that were already
adequate in their dietary intakes. As calcium and zinc intakes
were low, our advice is to increase the intake of meat, fish, dairy
products, legumes and nuts, foods rich in these nutrients before
prescribing supplements. This advice also increases the child’s
energy intake. Additional counselling is necessary to ensure
that the energy intake is appropriate when the children are in an
active disease state and are often anorexic.
In summary, we have shown that poor nutritional status is
evident in children with systemic or polyarticular disease, with
particular concern due to inadequate dietary intakes of energy,
calcium and zinc. This indicates that nutritional assessment is
essential in the management of patients with JCA, in particular
those with systemic or polyarticular disease.
NOTE ADDED tN PROOF
Following acceptance of this paper a further report has been
published. Bacon M. C., White P. H., Raiten 0.J. eta/. Nutritional
A. L. Mortensen et a/.
status and growth in junvenile rheumatoid arthritis. Semin.
Arthritis. Rheum. 1990: 20; 97-106.Poor calorie and zinc intake
were also commonly found, particularly in the systemic onset
group.
ACKNOWLEDGEMENTS
The authors thank Donna Waters for statistical assistance and
Drs Macauley and Champion for referral of patients. This work
was supported by grants from The Children’s Hospital Fund and
George Weston Food Pty Limited.
REFERENCES
1 Cassidy J. T.. Nelson A. M. The frequency of juvenile arthritis. J.
Rheurnatol. 1988; 15: 535-6.
2 Ansell 8. M. Chronic arthritis in childhood. Ann. Rheum. Dis. 1978;37:
107-20.
3 Still G. F. On a form of chronic joint disease in children. Med. Chir.
Trans. 1897; 80: 47. (Reprinted Arch. Dis. Child. 1941; 16: 156-65)
4 Ansell 8. M., Bywaters E. G. L.Alternate-day corticosteroidtherapy in
juvenile chronic polyarthritis. J. Rheumatol. 1974; 1: 176-86.
5 Allen-R.,Jimenez M.. Cowell C. Insulin-likegrowth factor and growth
hormone secretion in juvenile chronic arthritis. Ann. Rheum. 1990 (in
press).
6 Portinsson A,. Akesson A,, Svantesson H., Akesson B. Dietary
assessment in children with juvenile chronic arthritis. J. Hum. Nutf.
Diet. 1988; 1: 87 -94.
7 Lovell D. J.. Gregg D.. Huebi J.. Levinson J. E. Nutritional status in
juvenile rheumatoid arthritis-An interim report. Conference on
Rheumatic Diseases of Childhood, Park City. Ufah. 1986.
8 Hintz R. L.. Suskind R.. Amatayakul K.,Thanngkul 0.. Olson R. Plasma
somatomedin and growth hormone values in children with protein-
calorie malnutrition. J. Pediatr. 1978; 92: 153-6.
9 Cameron N. The methods of auxological anthropometry. In: Falkner
F.,Tanner J. M.. eds. Human Growth, Vol2. Plenum Press, New York.
1978; 35-90.
10 National Centre for Health Statistics growth charts. United States
Public Health Service. Rockville MD 1976. Measuring Change in
Nutritional Status. World Health Organization, Geneva. 1983.
11 Durnin J. V. G. A., Rahaman M. M. The assessment of the amount of
fat in the human body from skinfold thickness. Brit. J. Nutr. 1967; 21:
681-9.
12 Brook C. G. D.Determination of body composition of children from
skinfold measurements. Arch. Dis. Child. 1971; 46: 182-4.
13 Frisancho A. R. New norms of upper limb fat and muscle areas for
assessment of nutritional status. Amer. J. Clin. Nutr. 1981; 34: 2540-5.
14 National Health and Medical Research Council. Recommended
Dietary lntakes for Use in Australia.Australian Government Report of
Publishing Service, Canberra 1987.
15 World Health Organization. Energy and protein requirements. World
Health Organization Technical Report Series 724. Geneva. 1985.
16 Johansson U.. Portinsson S.. Akesson A,, Svantesson H., Ockerman
P. A., Akesson 8. Nutritional status in girls with juvenile chronic
arthritis. Hum. Nutr. Clin. Nutr. 1986; 40C: 57-67.
17 Hyams J. S., Carey D. E. Corticosteroids and growth. J. Pediatr. 1988;
113: 249-54.
18 Nelson M., Black A. E.. Morris J. A.. Cole T. J. Between- and within-
subject variation in nutrient intake from infancy to old age: estimating
the number of days required to rank dietary intakes with desired
precision. Amer. J. Clin. Nutr. 1989; 50: 155-67.
39 Deparlment of Community Services and Health. Nafional Dietary
Survey of School Children (aged 10- 15 years): 1985. No. 2. Nutrient
infakes.Australian Government Publishing Service, Canberra. 1989.
20 Nestle Nutrition Workshop Series, Vol. 14. J. C. Waterlow (ed.). Linear
Growth Retardation, in Less Developed Countries. Raven Press, New
York. 1988.
21 Bigaouelte J.. Timchalk M.. Kremer J. Nutritional adequacy of diet and
supplements in patients with rheumatoid arthritis who take medi-
cations. J. Amer. Diet. Assoc. 1987; a?: 1687-8.
22 Wahlquist M.. Huang S. Use and Abuse of Vitamins. Sun Books.
Melbourne, 1983; 86-7.