European Journal of Clinical Nutrition (2002) 56, 947–951 ß 2002 Nature Publishing Group All rights reserved 0954–3007/02 $25.00 www.nature.com/ejcn ORIGINAL COMMUNICATION Determinants of weight and length of Indonesian neonates

S Muslimatun1,2, MK Schmidt1,2, CE West2,3*, W Schultink4, R Gross5 and JGAJ Hautvast2

1SEAMEO TROPMED Regional Center for Community Nutrition, Jakarta, Indonesia; 2Division of and Epidemiology, Wageningen University, The Netherlands; 3Department of , University Medical Center Nijmegen, The Netherlands; 4Micronutrients, UNICEF, New York, USA; and 5German Agency for Technical Cooperation (GTZ), Eschborn, Germany

Objective: To investigate the determinants of neonatal weight and length. Design: From 16 – 20 week of , 366 mothers of the neonates had participated in the community-based study to investigate the effect of weekly supplementation during pregnancy with and A on infant growth. Women from five villages were allocated randomly to receive two tablets each containing 60 mg iron as ferrous sulphate and 250 mg folic acid (n ¼ 121) or two tablets each containing 2400 RE in addition to the same amount of ferrous sulphate and folic acid (n ¼ 122). A third (‘daily’) group (n ¼ 123) participating in the national iron supplementation programme was recruited from four neighbouring villages. Results: Neonatal weight and length did not differ between the two weekly groups and between the weekly iron group and the ‘daily’ group. Iron and vitamin A status during pregnancy did not influence neonatal weight and length significantly. Boys were 100 g heavier and 0.53 cm longer than girls (P < 0.05). First born neonates were lighter (P < 0.01) and tended to be shorter (P ¼ 0.070) than neonates of higher birth order. Maternal age and education as well as other socioeconomic determinants were not associated with neonatal weight and length. Neonatal weight was 32% explained by , maternal weight, postnatal measurement, gender and parity, while neonatal length was 28% explained by gestational age, maternal weight, postnatal measurement, gender and maternal height. Conclusions: Gestational age, maternal weight at second trimester and infant gender were the main predictors of neonatal weight and length. Sponsorship: The study was supported by The Netherlands Organization for Scientific Research-Netherlands Foundation for the Advancement of Tropical Research (NWO-WOTRO; WV 93-280) and Neys-van Hoogstraten Foundation (IN 114), The Nether- lands, and German Agency for Technical Cooperation (GTZ)=South East Asian Ministers of Education Organization (SEAMEO), Indonesia. European Journal of Clinical Nutrition (2002) 56, 947 – 951. doi:10.1038/sj.ejcn.1601439

Keywords: Indonesia; iron; neonatal length; neonatal weight; socioeconomic; supplementation; vitamin A

Introduction tries, one-sixth of babies have a low ( < 2500 g) Barker (1994) has hypothesized that foetal undernutrition (de Onis et al, 1998a, b). Weight and length of newborn leads to various adult chronic diseases. In developing coun- infants are influenced to a certain extent by maternal nutri- tional status both before and during pregnancy (Kramer, 1987), while the effect of socioeconomic condition on birth weight is mediated through effects on the mother (Kramer, 1998; Tuntiseranee et al, 1999). *Correspondence: CE West, Division of Human Nutrition and Low weight gain particularly during second trimester Epidemiology, Wageningen University, PO Box 8129, 6700 EV Wageningen, The Netherlands. elevates the risk of intrauterine growth retardation (Abrams E-mail: [email protected] & Selvin, 1995; Strauss & Dietz, 1999; Abrams et al, 2000). Guarantors: W Schultink and CE West. However, efforts to control foetal nutrition through Contributors: The study was conceived by WS, RG, JGAJH, SM and improvement of maternal dietary intake have given contra- MKS. The latter two also carried out the study and data analysis. All the investigators were involved in data interpretation and writing the dictory results. In the Gambia, energy and protein supple- report. mentation resulted in an increased birth weight only when Weight and length of Indonesian neonates S Muslimatun et al 948 provided to pregnant women during a season of hunger and group participating in the national iron supplementation intense agriculture work (Prentice et al, 1987). Provision of programme was recruited from neighbouring villages at the additional energy during the last trimester to mothers in same time. According to government policy, pregnant Madura in East Java did not improve the birth weight of their women receive 90 – 120 tablets containing iron plus folic infants (Kardjati et al, 1988). acid throughout pregnancy distributed through medical Although many pregnant women, particularly in devel- services. Therefore it was regarded as unethical to have a oping countries, are deficient in several micronutrients, iron placebo group in which women did not receive iron supple- deficiency is the most widespread problem. Risk of preterm ments. The median number of iron tablets consumed by delivery and are associated with iron women in the ‘daily’ iron group was 50. deficiency anaemia during pregnancy (Scholl & Reilly, 2000). The addition of vitamin A to routine iron supplemen- tation has been shown to increase haemoglobin concentra- Socioeconomic and pregnancy history indicators tions during pregnancy (Suharno et al, 1993, Muslimatun Demographic and socioeconomic characteristics and preg- et al, 2001). nancy history were assessed through interview using a pre- In addition to the intake of energy and nutrients, socio- coded questionnaire at enrolment following the recommen- economic factors are associated with nutritional status dations of Gross et al (1997). Socioeconomic variables during pregnancy and determine, in part, neonatal weight included were occupation, educational level, household pos- and length. In planning effective supplementation, it is of sessions and hygiene facilities. Parity and the number of particular importance to identify these crucial socioeco- miscarriages were also reported. Village estimated nomic factors. As part of our randomized control trial to the gestation time by palpation and from the date of last investigate the effect of weekly iron and vitamin A supple- menstruation period and the date was later verified against mentation during pregnancy on infant growth in the first the date of actual . year of life, we also collected data to evaluate how prenatal factors influence birth weight and length. In this paper, our first objective was to investigate the association between Anthropometry assessment socioeconomic and maternal nutritional factors while our Maternal anthropometric measurements were assessed at second objective was to investigate the determinants of enrolment (  18 weeks pregnancy) and near term (35 neonatal weight and length. Results on the effect of iron weeks pregnancy). Body weight was measured to the nearest and vitamin A supplementation on the iron and vitamin A 0.1 kg using a UNICEF electronic weighing scale (SECA 890, status near term (Muslimatun et al, 2001) has been reported Hamburg, Germany), mid-upper-arm circumference (MUAC) elsewhere. was measured to the nearest 0.1 cm using a plastic measuring tape, and height was measured to the nearest 0.1 cm using a standing height measurement ‘microtoise’ (UNICEF, Copen- Materials and methods hagen, Denmark). Neonatal weight and length were mea- Description of the study population sured during a postnatal home visit. Neonatal weight was The study was conducted in nine out of 19 rural villages in measured to the nearest 50 g using a baby weighing scale Leuwiliang sub-district, West Java, Indonesia from November (Misaki, Japan) and length was measured to the nearest 1997 until November 1999. Three villages were in the service 0.1 cm using a wooden length board (Gross et al, 1997) by area of the main health centre and six other villages were in two of the authors (SM and MKS). two other smaller health centres, which were under super- vision of the main health centre. Leuwiliang, 80 km south of Jakarta, has an area of 158 km2 and a population of about 140 Iron and vitamin A status assessments in serum 000. The maximum distance between study villages was On the same day as anthropometric assessment was carried 20 km. The area was rural and hilly. Only the main road out, venous blood samples (  5 ml) were collected in a tube had asphalt and the remaining roads were made from stone without anticoagulant between 09:00 and 12:00 h. Haemo- or soil. Public transportation was available throughout the globin was determined using the cyanmethaemoglobin day. method (Merck test 3317; Merck, Darmstadt, Germany) at Study design has been described earlier (Muslimatun et al, the laboratory of the Nutrition Research and Development 2001). In brief, women who were 16 – 20 weeks pregnant, Center, Bogor. Serum was analyzed by enzyme immu- aged 17 – 35 y, and parity < 6 were assigned randomly to two noassay using a commercial kit (IMX System, Abbott, Abbott weekly groups on an individual basis. They were supplemen- Park, IL, USA) at the SEAMEO TROPMED laboratory, Jakarta ted once weekly from enrolment until delivery with two and duplicate analysis were performed on 12% of the sam- tablets each containing 60 mg iron as ferrous sulphate and ples. Serum retinol was analyzed using high-pressure liquid 250 mg folic acid or with two tablets each containing 2400 chromatography (HPLC) at the Division of Human Nutrition retinol equivalents (RE) vitamin A in addition to the same and Epidemiology, Wageningen University. Ten percent of amount of ferrous sulphate and folic acid. A third (‘daily’) the analyses were carried out in duplicate. All analysis had

European Journal of Clinical Nutrition Weight and length of Indonesian neonates S Muslimatun et al 949 been described in more detail earlier (Muslimatun et al, Table 1 General characteristics of women at the time of enrolmenta 2001) Maternal characteristics n Weight (kg) Height (cm) MUAC (cm)

Age: Data analysis 17 – 21 y 122 48.1Æ 5.1 149.1Æ 4.2 24.1Æ 2.1** 22 – 26 y 126 49.2Æ 6.6 149.1Æ 4.9 24.7Æ 2.3 The normality of data distribution was checked using the > 26 y 118 49.4Æ 7.4 149.4Æ 5.3 25.1Æ 2.6 Kolmogorov – Smirnov test. Normally distributed data are Educational level: reported as mean and s.d. or s.e.m. Serum ferritin concentra- Primary school 269 48.4Æ 6.3** 149.1Æ 4.8 24.4Æ 2.3* tions were not normally distributed, therefore these data Secondary school 97 50.4Æ 6.6 149.7Æ 4.7 25.2Æ 2.3 Husband’s occupation:b were logarithmically transformed and reported as geometric Daily paid worker 165 48.4Æ 6.3 149.1Æ 4.9 24.5Æ 2.3 mean and 95% confidence interval. ANOVA and t-test were Monthly paid worker 128 49.5Æ 6.1 149.4Æ 4.8 24.9Æ 2.4 used to examine the relationship between maternal nutri- Trader 73 49.0Æ 7.2 149.3Æ 4.8 24.5Æ 2.6 tional status and socioeconomic and other individual char- Land: Yes 99 49.2Æ 6.9 149.1Æ 5.0 25.1Æ 2.7* acteristics and to compare the weight and length of neonates No 267 48.8Æ 6.2 149.2Æ 4.7 24.4Æ 2.2 born to mothers varying in nutritional status. Analysis by Television: stepwise multiple linear regression was used to examine the Yes 210 49.2Æ 6.8 149.2Æ 4.8 24.9Æ 2.6* determinant factors of neonatal weight and length. Only No 156 48.5Æ 5.9 149.3Æ 4.8 24.3Æ 1.9 Set of chairs: variables that had significant association in univariate ana- Yes 235 49.5Æ 6.6* 149.4Æ 4.8 24.9Æ 2.4** lysis were included in the model. MUAC was not included in No 131 47.9Æ 5.9 148.8Æ 4.8 24.1Æ 2.2 the model because it was highly correlated with weight Radio and=or cassette player: Æ Æ Æ (r ¼ 0.812). The SPSS software package (Windows version Yes 258 49.2 6.4 149.3 4.8 24.8 2.5* No 108 48.3Æ 6.4 148.9Æ 4.7 24.1Æ 2.0 7.5.2, SPSS Inc., Chicago, IL, USA) was used for all statistical Bicycle: analyses and a P value of < 0.05 was considered as signifi- Yes 112 49.9Æ 7.1* 149.1Æ 4.7 25.2Æ 2.5** cant. No 254 48.4Æ 6.0 149.3Æ 4.8 24.3Æ 2.3 Place of defecation: Latrine 169 49.5Æ 6.5* 149.6Æ 4.8 24.9Æ 2.5* River=pond 197 48.2Æ 6.2 148.8Æ 4.8 24.3Æ 2.2 Results Sources of drinking water: All women were Moslems, 95% originated from one single Well 290 48.9Æ 6.0 149.0Æ 4.8 24.7Æ 2.3 Spring water 76 48.8Æ 7.8 149.8Æ 4.7 24.3Æ 2.7 ethnic group (Sunda), and 95% did not earn any money. Parity: Pregnant women with primary school education had signifi- 0 110 48.2Æ 5.1 149.0Æ 4.4 24.0Æ 2.0** cantly lower body weight and MUAC than their peers with 1 – 2 181 49.3Æ 6.8 149.4Æ 5.1 24.8Æ 2.4 secondary school education (Table 1). Lower age, primiparity 3 – 5 75 49.0Æ 7.3 149.1Æ 4.6 25.0Æ 2.5 and owning land, a set of chairs, television, radio and=or aMeanÆ s.d. cassette player were all associated with a higher MUAC. bDaily paid workers include farmers, labourers, drivers; monthly paid workers Height was not associated with age, education, occupation, include teachers, government and private employees. possession of household goods, environmental conditions *P < 0.05; **P < 0.01; ANOVA controlling for time of measurement during pregnancy. and parity. As expected, primiparous women (30% of the total) were significantly younger than multiparous women (19.9Æ 2.2 vs 26.1Æ 4.1 y, respectively; P < 0.01; meanÆ s.d.). The proportion of subjects who had miscarried was 11% sured later than those in the weekly groups (6.9Æ 3.7 vs (n ¼ 36). Parity and age were not associated with the 5.2Æ 3.3 days; P < 0.01). The period between birth and the number of miscarriages. postnatal home visit varied largely due to the difficulties in Out of the 366 pregnant women initially enrolled, 25 communication and reaching the subjects. The length of dropped out during pregnancy, 10 gave birth to a stillborn four neonates was not measured. Neonatal weight child, three had twins (only one survived), and two moved. (3094Æ 440 g) and length (49.1Æ 2.0 cm) did not differ The weight, height and MUAC of mother having stillborn or between the two weekly groups and between the weekly twin babies did not differ from mothers having live born iron group and the ‘daily’ group (P > 0.05, after controlling babies (data not shown). The majority of births took place at for age at measurement and gender, Table 2). Within the home (89%) and with the assistance of traditional birth ‘daily’ group, neonates from mothers who consumed  50 attendants (80%). Only a few births were assisted by mid- iron tablets during pregnancy had similar weights compared wives (15%) or medical doctors (3%). to those who consumed < 50 iron tablets. Also, iron and Weight and length were reported from 296 (91%) neo- vitamin A status during pregnancy did not significantly nates based on measurement within 15 days of birth. Mater- influence neonatal weight and length (data not shown). nal anthropometric parameters at enrolment of these Boys were 100 g heavier and 0.53 cm longer than girls neonates did not differ significantly from the other neonates (P < 0.05, controlling for age at measurement). The duration (data not shown). Neonates in the ‘daily’ group were mea- of pregnancy of women with anaemia (haemoglobin con-

European Journal of Clinical Nutrition Weight and length of Indonesian neonates S Muslimatun et al 950 Table 2 Factors associated with neonatal weight and lengtha Table 3 Multiple regression analysis on the determinants of neonatal weight and length Neonatal Neonatal n weight (g) n length (cm) Standardized coefficient Adjusted R2

Treatment group: Neonatal weight (n ¼ 296) 0.321 Weekly vitamin A and iron 103 3094Æ 455 101 49.1Æ 2.1 Gestational age 0.367 Weekly iron 102 3096Æ 396 101 49.1Æ 2.0 Maternal weight at enrolment 0.347 ‘Daily’ iron 91 3089Æ 475 90 49.1Æ 2.0 Age at measurement 0.231 Gender: Gender (0 ¼ boy, 1 ¼ girl) 7 0.158 Boys 142 3146Æ 436* 140 49.4Æ 2.1* Parity 0.147 Girls 154 3045Æ 441 152 48.8Æ 2.0 Neonatal length (n ¼ 292) 0.276 Age at measurement: Gestational age 0.321 0 – 3 days 71 3028Æ 333* 70 48.6Æ 1.7** Age at measurement 0.310 3 – 7 days 131 3059Æ 424 131 48.9Æ 2.1 Maternal weight at enrolment 0.220 7 – 15 days 94 3191Æ 515 91 49.8Æ 2.0 Gender (0 ¼ boy, 1 ¼ girl) 7 0.176 Gestational age: Maternal height at enrolment 0.125 < 37 weeks 58 2835Æ 496** 91 48.1Æ 2.2**  37 weeks 238 3155Æ 403 201 49.3Æ 1.9 Maternal weight: < 45.9 kg 100 2968Æ 392** 99 48.7Æ 1.9** Maternal age and education as well as other socioeco- 45.9 – 50.7 kg 101 3038Æ 427 98 48.9Æ 2.1  50.7 kg 95 3285Æ 443 95 49.7Æ 1.9 nomic determinants were not associated with neonatal Maternal MUAC: weight and length. From multiple regression analysis, gesta- < 23.4 cm 97 2968Æ 417** 94 48.7Æ 2.0* tional age, maternal weight, postnatal measurement and 23.4 – 25.5 cm 106 3096Æ 408 105 49.1Æ 2.0 gender determined both neonatal weight and length (Table  25.5 cm 93 3221Æ 467 93 49.5Æ 2.0 Maternal height: 3). In additional, parity determined neonatal weight while < 147.4 cm 99 2992Æ 406** 98 48.6Æ 1.8** maternal height determined neonatal length. Of the 296 147.7 – 150.9 cm 95 3093Æ 410 92 49.1Æ 1.7 neonates, data on maternal weight gain during pregnancy  Æ Æ 150.9 cm 102 3193 480 102 49.6 2.3 were available from 226 (76%) subjects. On average, women Maternal age: 17 – 21 y 98 3030Æ 432 94 49.1Æ 1.9 gained 5.0Æ 2.5 kg (0.29Æ 0.14 kg=week) from  18 until 35 22 – 26 y 107 3128Æ 440 107 49.1Æ 2.1 weeks of pregnancy. When we included this variable into the 26 – 35 y 91 3121Æ 448 91 49.1Æ 2.1 model, 37% of neonatal weight and 30% of neonatal length Maternal educational level: were explained. Determinants of neonatal weight did not Primary school 216 3099Æ 433 214 49.1Æ 1.9 Secondary school 80 3079Æ 464 78 49.0Æ 2.1 change, but gestational age and maternal height were no Paternal occupation:b longer significant determinants of neonatal length (data not Daily paid worker 140 3127Æ 439 139 49.2Æ 2.0 shown). Monthly paid worker 98 3071Æ 409 97 49.1Æ 1.9 Trader 58 3050Æ 494 56 49.0Æ 2.2 Maternal parity: 0 84 2920Æ 392** 80 48.6Æ 1.8 Discussion 1 – 2 148 3156Æ 436 148 49.2Æ 2.1 In this study, we showed that predictors of neonatal weight 3 – 5643177Æ 454 64 49.3Æ 2.1 and length were gestational age, maternal weight at second aMeanÆ s.d. trimester and infant gender. Weekly supplementation of bDaily paid workers include farmers, labourers, drivers; monthly paid workers vitamin A and iron during pregnancy did not influence include teachers, government and private employees. neonatal weight and length. Socioeconomic variables were *P < 0.05; **P < 0.01; ANOVA, controlling for age at measurement. associated with maternal nutritional status but not with neonatal weight and length. The influence of education and possession of household goods on maternal weight centrations < 110 g=l) at enrolment was shorter than those reflects the role of socioeconomic and environmental con- without anaemia (37.4Æ 2.7 vs 38.0Æ 2.8 weeks, P < 0.05). ditions. We pooled neonatal weight and length from the three Maternal weight prior to pregnancy is an important treatment groups to investigate factors associated with neo- determinant of foetal outcomes, such as low birth weight, natal weight and length. Compared to the lowest tertile, intrauterine growth retardation and (Kramer, neonates from mothers in the highest tertile of maternal 1987; WHO, 1995). Also, weight gain, particularly during the weight, MUAC, and height at  18 weeks pregnant were second trimester, has been known to influence birth weight heavier and longer (P < 0.01, controlling for postnatal mea- positively (Abrams & Selvin, 1995; Strauss & Dietz, 1999; surement and gender, Table 2). Preterm neonates (gestational Abrams et al, 2000). In practice, weight gain monitoring or age < 37 completed weeks) were significantly (P < 0.001) assessment of total weight gain during pregnancy is imple- lighter and shorter than term neonates. First-born neonates mented only rarely in developing countries where early were lighter (P < 0.01) and tended to be shorter (P ¼ 0.07) is very weak or non-existent. Only in very than neonates of higher birth order. advanced stages of pregnancy do women attend health

European Journal of Clinical Nutrition Weight and length of Indonesian neonates S Muslimatun et al

References 951 services. It is recommended that women should have a Abrams B & Selvin S (1995): Maternal weight gain pattern and birth weight gain of at least 6 kg (1 kg per month) during the weight. Obstet. Gynecol. 86, 163 – 169. second and third trimesters (Krasovec, 1991). On average, Abrams B, Altman SL & Pickett KE (2000): Pregnancy weight gain: our population had met this recommendation. still controversial. Am. J. Clin. Nutr. 71(Suppl), 1233S – 1241S. Achadi EL, Hansell MJ, Sloan NL & Anderson MA (1995): Women’s We did not find an association between maternal iron nutritional status, iron consumption and weight gain during status on neonatal weight and length, but anaemia was pregnancy in relation to neonatal weight and length in West associated with shorter gestation. However, the difference Java, Indonesia. Int. J. Gynecol. Obstet. 48(Suppl), S103 – S119. of 0.6 weeks might not be biologically significant. In contrast Allen LH (2000): Anemia and iron deficiency: effects on pregnancy outcome. Am. J. Clin. Nutr. 71(Suppl), 1280S – 1284S. to findings from observational studies (Allen, 2000; Scholl & Barker DJP (1994): Mothers, Babies, and Disease in Later Life. London: Reilly, 2000), controlled trials on iron supplementation do BMJ Publishing Group. not provide evidence for benefits on preventing intrauterine de Onis M, Blo¨ssner M & Villar J (1998a): Levels and patterns of growth retardation (de Onis et al, 1998a,b). Preziosi et al intrauterine growth retardation in developing countries. Eur. J. Clin. Nutr. 52(S1), S5 – S15. (1997) showed that infants of iron supplemented mothers de Onis M, Villar J & Gulmezoglu M (1998b): Nutritional interven- were longer, but not heavier, at birth than those of non- tions to prevent intrauterine growth retardation: evidence from supplemented mothers. randomized controlled trials. Eur. J. Clin. Nutr. 52(S1), S83 – S93. Poor maternal vitamin A status was found to be associated Gross R, Kielmann A, Korte R, Schoeneberger H & Schultink W (1997): Guidelines for nutrition baseline surveys in communities. with prematurity and intrauterine growth retardation in a Bangkok: SEAMEO-TROPMED and Deutsche Gesellschaft fu¨r Tech- low-income group, but not in a high-income group, of an nische Zusammenarbeit (GTZ). Indian population in Baroda (Shah & Rajalakshmi, 1984). Howells DH, Haste F, Rosenberg D, Brown IRF & Brooke OG (1986): The mean neonatal weight and maternal serum retinol Investigation of vitamin A nutrition in pregnant British Asians and their infants. Hum. Nutr. Clin. Nutr. 40C,43– 50. concentrations of subjects in our study were comparable to Kardjati S, Kusin JA & de With C (1988): Energy supplementation in those from high-income groups, which explains why mater- the last trimester of pregnancy in East Java: I. Effect on birth- nal vitamin A status and neonatal weight and length were weight. Br. J. Obstet. Gynaecol. 95, 783 – 794. not associated. An earlier study revealed that, despite an Kramer MS (1987): Determinants of low birth weight: methodologi- cal assessment and meta-analysis. Bull. WHO 65,663– 737. improvement of serum retinol concentrations after daily Kramer MS (1998): Socioeconomic determinants of intrauterine supplementation with 8000 IU vitamin A among Indo-Pakis- growth retardation. Eur. J. Clin. Nutr. 52(S1), S29 – S33. tani women living in South London, the anthropometric Krasovec K (1991): Part I. Weight gain in pregnancy. In: Maternal indices of the babies were not affected (Howells et al, 1986). Nutrition and Pregnancy Outcomes. Anthropometric Assessment,ed.K Krasovec & MA Anderson, pp 15 – 55. Washington, DC: PAHO. In West Java, socioeconomic factors have been found to Muslimatun S, Schmidt MK, Schultink W, West CE, Hautvast JGAJ, be associated with nutritional status of pregnant women but Gross R & Muhilal (2001): Weekly supplementation with iron and not with the neonatal weight and length (Achadi et al, 1995). vitamin A during pregnancy increases hemoglobin concentration In Thailand (Tuntiseranee et al, 1999), family income was the but decreases serum ferritin concentration in Indonesian pregnant women. J. Nutr. 131,85– 90. only factor found to be strongly associated with birth weight. Prentice AM, Cole TJ, Foord FA, Lamb WH & Whitehead RG (1987): In our population, maternal age was not associated with Increased birthweight after prenatal dietary supplementation of birth weight as shown also by other studies (Abrams & rural African women. Am. J. Clin. Nutr. 46, 912 – 925. Preziosi P, Prual A, Galan P, Daouda H, Buoreima H & Hercberg S Selvin, 1995; Achadi et al, 1995). Primiparous women, who (1997): Effect of iron supplementation on the iron status of were younger than multiparous women, gave birth to babies pregnant women: consequences for newborns. Am. J. Clin. Nutr. with lower birth weight (Abrams & Selvin, 1995). 66, 1178 – 1182. In conclusion, gestational age, maternal weight at second Scholl TO & Reilly T (2000): Anemia, iron and pregnancy outcome. J. Nutr. 130, 443S – 447S. trimester and infant gender were the main predictors of Shah RS & Rajalakshmi R (1984): Vitamin A status of the newborn in neonatal weight and length in our population. The impor- relation to gestational age, body weight, and maternal nutritional tance of socioeconomic and environmental conditions on status. Am. J. Clin. Nutr. 40, 794 – 800. maternal nutritional status suggests that effort to improve Strauss RS & Dietz WH (1999): Low maternal weight gain in the second or third trimester increases the risk for intrauterine growth maternal nutritional status before and during pregnancy retardation. J. Nutr. 129, 988 – 993. should be encouraged. Although iron and vitamin A supple- Suharno D, West CE, Muhilal, Karyadi D & Hautvast JGAJ (1993): mentation during pregnancy did not affect neonatal weight Supplementation with vitamin A and iron for nutritional anaemia and length, it may influence subsequent nutritional status in pregnant women in West Java, Indonesia. Lancet 342, 1325 – 1328. and growth, as well as mental and motor development. Tuntiseranee P, Olsen J, Chongsuvivatwong V & Limbutara S (1999): Socioeconomic and work related determinants of pregnancy out- come in southern Thailand. J. Epidemiol. Community Health 53, Acknowledgements 624 – 629. WHO (1995): Maternal anthropometry and pregnancy outcomes. A We gratefully acknowledged the participation and co-opera- WHO collaborative study. Bull. WHO 73(Suppl), 1 – 98. tion of pregnant women, heads of villages and health centres in Leuwiliang, Bogor, Indonesia. The comment on the manuscript provided by Dr AA Kielmann is highly appre- ciated.

European Journal of Clinical Nutrition