Nutritional Status and Linear Growth of Indonesian Infants in West Java Are
Determined More by Prenatal Environment than by Postnatal Factors1
Marjanka K. Schmidt,* Siti Muslimatun,* Clive E. West,* **2 Werner Schultink,
Rainer Gross and Joseph G. A. J. Hautvast*
*Division of Human Nutrition and Epidemiology, Wageningen University, The Netherlands; SEAMEO
TROPMED Regional Center for Community Nutrition, University of Indonesia, Jakarta, Indonesia;
**Department of Gastroenterology, University Medical Center Nijmegen, The Netherlands; UNICEF,
New York, NY; and German Agency for Technical Cooperation, Eschborn, Germany.
KEY WORDS:
infants
growth
nutritional status
In Indonesia, many infants and young children have an inadequate nutritional status as reflected by the high prevalence of
underweight and stunting (1,2). Stunting indicates a public
health problem because of its association with an increased risk of
morbidity and mortality, delayed motor and mental development,
and reduced physical capacity (1,3). Most growth faltering, resulting in underweight and stunting, occurs from birth until 2 y
of age but may already start during the gestation period (3).
Prenatal factors, such as maternal nutrition, determine fetal
growth during the gestation period, and thus infant weight and
length at birth. Infant birth weight and length are determinants of future nutritional status as was shown in Bangladeshi
infants who experienced very little catch-up growth after birth
(4). In Indonesian infants, birth weight and length have also
been found to be important determinants of infant growth (5).
During infancy, inadequate intakes of nutrients and frequent
or prolonged episodes of infection may exacerbate the effects of
feeding
morbidity
determinants
1
Supported by the Netherlands Organization for Scientific Research-Netherlands Foundation for the Advancement of Tropical Research and the Neys-van
Hoogstraten Foundation, The Netherlands and the German Agency for Technical
Cooperation/South East Asian Ministers of Education Organization, Tropical
Medicine (SEAMEO TROPMED), Indonesia.
2
To whom correspondence should be addressed.
E-mail: Clive.West@staff.nutepi.wau.nl.
ABSTRACT One of the health problems in Indonesia is the high prevalence of stunting in infants. Determinants
and specifically the relative contribution of prenatal and postnatal factors to growth and nutritional status of
Indonesian infants were investigated. Newborn infants, from women recruited at 18 wk of pregnancy from 9 rural
villages in West Java, Indonesia, were followed until 1215 mo of age. Weight, length, morbidity, breast-feeding
and food intake were assessed monthly. Determinants of length and weight increase and nutritional status reflected
by Z-scores were evaluated using multiple linear regression. Neonatal weight (3.2 0.5 kg) and length (49.7 2.2
cm) were reasonable. However, growth started to falter at 6 7 mo of age, resulting in prevalences of 24% stunting
and 32% underweight at 12 mo of age. The multiple regression models explained 19 41% of the variation in
growth and nutritional status of infants. Neonatal weight ( 0.285) and length ( 0.492) were the strongest
positive predictors of weight-for-age and height-for-age Z-scores, respectively. Fever was negatively associated
with weight increase ( 0.144) and weight-for-age ( 0.142) and weight-for-height Z-scores ( 0.255)
but not with length increase or height-for-age Z-scores. Intake of complementary foods was positively associated
with increases in weight ( 0.190) and length ( 0.179) and nutritional status of infants ( 0.136 0.194). In
conclusion, in this rural population in West Java, neonatal weight and especially length, reflecting the prenatal
environment, are the most important predictors of infant nutritional status. J. Nutr. 132: 22022207, 2002.
born between February and November 1998. After birth, all eligible
infants were followed up monthly until at least 12 mo of age. Thereafter, infants whose mothers were still willing to participate continued to be measured into childhood until the end of the study period
(November 1999). Growth, morbidity, breast-feeding and food intake
were assessed monthly. In addition, mothers were interviewed at
enrollment for socioeconomic background and pregnancy history, at
delivery for infant food intake and other variables and at 13 mo after
delivery about the use of contraceptives and pregnancy status. Selected anthropometric measurements of women (14) were assessed at
enrollment, near term and 4 mo postpartum. The data presented in
this paper were collected in an intervention trial investigating the
effect of vitamin A and iron supplementation during gestation on
infant growth. The results of the intervention and details of the
study design have been published (14 17). Explanation of the study
was given to the women at enrollment and all women who participated gave written informed consent. The Ministry of Health Indonesia and the Ethical Committees of the Medical Faculty at the
University of Indonesia and Wageningen University approved the
research proposal.
Study population. Of 366 women recruited at 18 wk of pregnancy, 318 infants were eligible for follow-up after birth. Loss of
subjects occurred because of withdrawal from supplementation (n
18), moving outside the research area (n 15), stillbirth or twins
(n 13), or neonatal death (n 2) (14,16). During the follow-up
period from birth to 12 mo, 16 infants moved outside the research
area and 12 infants died. Neonatal weight or length (in mo 1 of life)
and gender ratio did not differ between infants who left the study and
infants who were followed up to 12 mo (n 290). All 318 infants
experienced at least one anthropometric assessment. Infants who
were measured 19 (n 46), 10 13 (n 100) or 14 19 (n 172)
times did not differ with respect to neonatal weight or length, weight
gain or length increase during the first 6 mo of life and gender ratio.
The mean number of measurements was 13.2 3.6.
Assessment of infants. Weight and length of infants were measured and mothers interviewed for morbidity and food intake of their
infants during their monthly visit to their neighborhood health post
(Posyandu). In principle, visits for each health post were planned on
the same date of each month. However, in practice, visits were made
within 1 wk of the planned date. We attempted to reach individual
mothers who were not able to attend by visiting their homes.
Anthropometry. Two pairs of trained field assistants measured
weight and length of infants using standardized methods (18). Weight
was measured to the nearest 50 g using a baby weighing scale (Misaki
baby scale, Japan), which was tared each time before use. Calibration
with a standard weight (5 kg) at regular intervals showed that all
scales were stable and precise. Recumbent length was measured to the
nearest 0.1 cm using a wooden length board (19). Infants were
measured with light clothing. Birth or neonatal weight and length
had been measured by two of the authors (M.K.S. and S.M.) during
a postnatal home visit (16). All field assistants had received training
and were supervised every month by the two authors mentioned
above. Independent measurements performed in random subsample of
5% of the infants showed a technical SD of 44 g and 0.70 cm for
weight and length, respectively (17).
Morbidity. At the time of the anthropometric assessment, the
same field assistants interviewed the caretaker (the mother in all but
2 cases) of the infant about symptoms of morbidity in the 14 d
preceding the visit. More specifically, mothers were asked to recall
the following: diarrhea, defined as 3 stools in 1 d that were more
liquid than usual; fever, mothers evaluation of infants body temperature above normal (hot to the touch; panas); running nose
(pilek), nasal discharge; cough, persistent coughing; difficulty
breathing, breathing with severe noise or wheezing or difficulty inhaling; ear discharge, fluid or pus draining from at least one ear; and
vomiting. Questions were asked in Bahasa Indonesia or in the local
language (Sundanese) if the mother did not understand the term.
First the interviewer asked the mother for any symptoms and then
asked her specifically for each category and for how many days the
infant suffered from that specific symptom. Episodes were defined as
a period of sickness separated by at least 3 d. Mortality of infants was
recorded at a home visit after death had been reported.
2203
SCHMIDT ET AL.
2204
RESULTS
General characteristics of the study population are shown
in Table 1. The number of boys and girls was not different;
27% of the infants were first-borns. Birth weight of infants was
reasonable (Table 2); 6% of the infants had low birth weight
TABLE 1
General characteristics of the cohort of Indonesian newborn
infants and their mothers1
Boy/girl, n
Birth order
Gestational age, wk
Maternal age at enrollment, y
Maternal height at enrollment, cm
Maternal weight at enrollment, kg
Maternal MUAC2 at enrollment, cm
Maternal education, %
Elementary school
Higher education
Occupation of father, %
Daily worker (laborer, farmer, driver)
Monthly worker (government or private employee,
retired)
Trader
Housing, %
Good housing (brick wall, tile roof and cemented
floor)
Poorer housing (wall, roof and/or floor made of
materials with less quality)
1 Data are means SD unless stated otherwise.
2 MUAC, mid-upper arm circumference.
151/167
1.5 1.3
37.8 2.6
24.3 4.6
149.1 4.9
48.9 6.6
24.6 2.4
74.8
25.2
46.5
34.0
19.5
67
33
TABLE 2
Weight, length and growth of Indonesian infants from
birth to 15 mo of age1
Age
72
h2
0 mo3
6 mo
12 mo
15 mo
68
3124
267
248
1285
Period
mo
06
612
1215
2634
223
1175
Weight
Length
kg
cm
3.0 0.3
3.2 0.5
7.3 0.9
8.2 0.9
8.7 0.9
48.5 1.7
49.7 2.2
65.5 2.4
71.7 2.4
74.4 2.5
Weight increase
Length increase
kg
cm
4.1 0.8
1.0 0.5
0.4 0.3
16.0 2.1
6.4 1.2
2.4 1.1
1 Mean SD.
2 Birth weight and length of only those infants who could be
maternal height and nutritional status (weight, mid-upper arm circumference) during pregnancy and infant neonatal weight and
length; postnatal factors such as duration of breast-feeding, food
practices and morbidity; and covariants such as gender, birth order
and age of the infants. All variables presented in Table 1 were
included and entered as dichotomous or continuous. Occupation of
the father was entered as dichotomous (0 daily worker, 1
monthly worker or trader) variable but was not a significant
determinant in the model. Mid-upper arm circumference of mothers
was not included in any model because it was highly correlated with
maternal weight (r 0.8).
The postnatal variables, prelacteal feeding (0,1), colostrum intake
(0,1) and number of months the infant was exclusively breast-fed,
were included in all models. The period of follow-up (age) was
calculated as the actual age at the 12-mo measurement minus the
actual age at the 0-mo measurement. This age variable was included
as a possible confounder in the models for weight and length increase.
When neonatal weight and length were both included in all models,
neonatal weight was a significant predictor in the models for weight
gain and weight-for-age Z-scores, neonatal length in the models for
length increase and height-for-age Z-scores, and both variables in the
model for weight-for-height Z-scores. Because neonatal weight and
length were highly correlated (r 0.8), models were rerun including
only the significant variable (neonatal weight or length, or both in
the case of weight-for-length). To limit the number of variables and
to avoid collinearity, we selected one variable for morbidity symptoms
and one for food groups to be included in the models. The selection
was based on a comparison of correlation coefficients of each morbidity or food variable with the dependent variable (data not shown).
Fever was most associated with growth and nutritional status of
infants and was therefore included in all models. Of the food groups,
fruit was most associated with weight gain, weight-for-age and heightfor-age Z-scores, biscuits with length increase and snacks with weightfor-height Z-scores; thus, these were included in the respective models. Statistical analysis was carried out with the SPSS software
package (Windows version 7.5.2. SPSS Chicago, IL).
2205
SCHMIDT ET AL.
2206
TABLE 3
Regression analysis to explain the influence of prenatal and postnatal determinants on growth
from 0 to 12 mo of age and nutritional status at 12 mo of age of Indonesian infants1
Increase in weight, 012 mo
227
226
SE
0.215
0.019
0.018
0.467
0.462
0.390
0.626
0.777
0.103
0.011
0.008
0.098
0.169
0.110
0.252
0.237
0.122
0.115
0.151
0.281
0.165
0.222
0.144
0.190
0.039
0.090
0.029
0.000
0.007
0.000
0.014
0.001
SE
0.428
0.086
0.246
0.023
0.091
0.196
0.083
0.000
1.123
1.929
0.433
0.235
0.403
0.054
0.251
0.256
0.435
0.000
0.000
0.000
2.102
0.606
0.179
0.001
Height-for-age Z-score
Weight-for-height Z-score
227
226
226
SE
SE
0.191
0.019
0.019
0.216
0.469
0.100
0.011
0.008
0.095
0.101
0.114
0.124
0.162
0.136
0.285
0.056
0.069
0.021
0.024
0.000
0.158
0.035
0.092
0.009
0.095
0.229
0.087
0.000
0.312
0.172
0.088
0.019
0.198
0.492
0.000
0.000
0.591
0.759
0.244
0.230
0.142
0.194
0.016
0.001
0.596
0.211
0.154
0.005
SE
0.021
0.007
0.195
0.002
0.730
0.147
0.042
0.207
0.043
0.022
0.481
0.459
0.117
0.001
0.001
0.059
0.957
0.229
0.255
0.000
0.634
0.288
0.136
0.029
1 Variables included in the models are explained in the methodology section, variables with dash () are excluded from the model; the percentage
of variation explained by the models (adjusted R2) was 25 and 41% of increases in weight and length from 0 to 12 mo, respectively, and 25% of
weight-for-age, 35% of height-for-age and 19% of weight-for-height Z-scores.
2 Unstandardized coefficient.
3 Standardized coefficient; this reflects the correlation of the dependent and independent variable without considering the unit of both variables,
thus making it possible to compare the strength of association with the dependent variable among the independent variables.
4 Fruit, biscuits and snacks: the number of intake interviews of the food group was reported per number of assessments (as explained in the
methodology section).
Weight-for-age Z-score
n
b2
ACKNOWLEDGMENTS
We are grateful to all participants, field assistants and health
personal of subdistrict Leuwiliang, Bogor for their contribution to this
project.
LITERATURE CITED
1. ACC/SCN. (2000) Fourth Report on the World Nutrition Situation.
Nutrition Throughout the Life Cycle. ACC/SCN in collaboration with IFPRI, Geneva, Switzerland.
2. World Health Organization. (2000) Global Database on Child Growth
and Malnutrition. WHO, Geneva, Switzerland.
3. Waterlow, J. C. & Schu rch, B. (1994) Causes and mechanisms of
linear growth retardation. Proceedings of an I/D/E/C/G Workshop Held in London
January 1518, 1993. Eur. J. Clin. Nutr. 48: S1S216.
4. Arifeen, S. E., Black, R. E., Caulfield, L. E., Antelman, G., Baqui, A. H.,
Nahar, Q., Alamgir, S. & Mahmud, H. (2000) Infant growth patterns in the slums
of Dhaka in relation to birth weight, intrauterine growth retardation, and prematurity. Am. J. Clin. Nutr. 72: 1010 1017.
5. Kolsteren, P. W., Kusin, J. A. & Kardjati, S. (1997) Pattern of linear
growth velocities of infants from birth to 12 months in Madura, Indonesia. Trop.
Med. Int. Health 2: 291301.
2207