Contents - Previous - Next

Results

Pre-pregnant BMI

Madurese women were short (average height 150 cm) and lean (average weight 42 kg). About 40% had a pre-pregnant weight of <42 kg and 13% a height <145 cm. Mean BMI increases with weight but there was no relation with height (Table 1). The great majority of women weighing <42 kg had a BMI <18.5 and can be considered as CED. Also the prevalence of CED did not differ by height groups and current height was thus not an indicator of poverty in this community (Table 2). The prevalence of severe, moderate and mild CED in this population was 3%, 9% and 26% respectively.

BMI by parity and birth interval are shown in Table 3. Unexpectedly, primipara had a lower mean BMI and a higher prevalence of CED. Such a pattern may be due to a number of primiparae being still adolescent. In the study population modern contraception is not common and parity will be closely related to age. Scrutiny of BMI by age and parity separately did not, however, reveal an age effect: primiparae of age 16-19 years, 20-22 years and 23-25 years had a mean BMI of 18.6, 18.9 and 18.0 respectively and no marked difference in incidence of low birth weight (<2500 g). The reason for primiparae being more undernourished remains to be elucidated.

BMI was not related to birth interval. Mean BMI for the groups with a birth interval of <18 months and >27 months was 19.4 and 19.3 respectively; the corresponding figures for the prevalence of CED were 32% and 34%. Apparently women of reproductive age were chronically undernourished to start with and it is more likely that the degree of undernutrition will influence reproductive performance rather than reproduction leading to maternal depletion.

BMI and outcome of pregnancy

Total weight gain in pregnancy was on average 6.6 kg. Compared with well-nourished British women (Taggart et al., 1967), skinfold thicknesses of Madurese women were much lower, viz. ~50% at the tricipital and bicipital sites and 15-20% at the subscapular and suprailiac sites (Kardjati et al., 1990).

In such a population, pre-pregnant maternal energy reserves may be an important determinant of fetal growth (Winikoff & Debrovner, 1981; Naeye, 1981; Abrams & Laros, 1986; Frenzen, Dimperio & Cruz, 1988; Mitchell & Lerner, 1989; NAS, 1991). There was indeed an upward gradient in mean birth weight by increasing pre-pregnant BMI (Table 4). Apparently there was a cut-off for risk of low birth weight. A very high incidence of low birth weight (22%) was observed only in the severe CED group but the incidence in the other CED groups (8-10%) was still two times higher than in the non-CED groups.

Table 2. Percentage of women with chronic energy deficiency (CED) by categories of pre-pregnant weight and height: percentages per cell

a = % BMI <16.0 (CED III).
b = % BMI 16.0-17.0 (CED 11).
c = % BMI 17.0-18.5 (CED 1).

Table 3. Pre-pregnant body mass index (BMI) by parity and birth interval: mean (SD)

Birth interval = number of months in between current and preceding birth.

Table 4. Birth weight by categories of pre-pregnant body mass index (BMI): mean, standard deviation and % low birth weight (LBW)

Birthweight

LBW <2500g. NS = not significant.

Cut-offs of pre-pregnant BMI <16 and <17 were specific but not sensitive indicators for low birth weight. The importance of pre pregnant BMI is illustrated by the relative risk for low birth weight which was 4 times higher at BMI <16 and twice as high at BMI <17. The attributable risk in the severely undernourished group (BMI <16) was 77% and in the other two categories 56% and 53% respectively (Table 5).

The full picture of the reproductive process can only be judged if fetal and maternal outcome of pregnancy are considered in conjunction. It is reasonable to assume that maternal body composition has returned to the non-pregnant state at 4-7 weeks post partum (Durnin & McKillop, 1988). The difference in weight and BMI before pregnancy and at 4-7 weeks post partum thus illustrates the net maternal weight changes in pregnancy. On average women gained a modest 0.7 kg but there was a clearly different response to reproductive stress if women were separated into prepregnant BMI categories (Table 6).

Table 5. Risk cut-offs of pre-pregnant body mass index (BMI) for low birth weight

Between parentheses = prevalence of each BMI category.
Se = sensitivity; Sp = specificity; RR = relative risk; AR = attributable risk.

Table 6. Net weight (kg) changes in pregnancy by Pre-pregnant body mass index (BMI) categories

Weight changes between BMI categories:
<16 vs 16-16.9: NS.
<17 vs 17-18.4: t = 2.56, P = 0.011.
<17 vs ³17: t=7.28, P<0.001.
<18.5 vs ³18.5: t = 10.38, P < 0.001.

CED women gained weight during pregnancy, viz. 2.4 kg in the severe and moderate CED groups and 1.7 kg in the mild CED group, irrespective of parity. Non-CED women did change weight in pregnancy if BMI was <22.5 but they lost 1.5 kg if they were fatter before pregnancy. The differences at a BMI cut-off of 17.0 and 18.5 were highly significant (P < 0.0001)

The first 12 months post partum

Genuine on-demand breastfeeding is still common in the study population, as elsewhere in rural Indonesia. The median duration of lactation in the EJPS community is 30 months. As mentioned earlier, the daily energy intake during lactation was ~200 kcal higher than during pregnancy. Also lactating women were at most moderately active. Due to the higher energy requirements for lactation compared with pregnancy (Steenbergen et al., 1991) and the low fat reserves built up during pregnancy, the energy deficit of nursing Madurese women may thus be even higher than during pregnancy.

Continue

Contents - Previous - Next