Advancing Sexual and Reproductive Health and Rights
International Family Planning Perspectives
Volume 31, Number 2, June 2005

Childhood Stunting Is Linked to the Pregnancy Intentions of Both Parents

In Bolivia, 12–35-month-old children whose mothers say their conception was unwanted or mistimed have a 30% higher risk of stunted growth than do those whose conception was intended.1 Data from the 1998 Demographic and Health Survey (DHS) also show that children in this age-group whose conception was reported as mistimed by both parents have a higher risk of stunting than do children whose conception was reported as intended by both parents (risk ratio, 3.8).

The DHS was conducted among a representative sample of 15–49-year-old women; it collected information on respondents' demographic characteristics, reproductive history, use of contraceptives and maternal health services, and lastborn child's height and age. In addition, women and their male partners were asked whether their last child had been wanted at the time of conception (intended), later (mistimed) or not at all (unwanted). Researchers calculated the prevalence of childhood stunting, which they defined as height-for-age more than two standard deviations below the 1978 reference median. They conducted bivariate and multivariate analyses to examine how stunting related to mothers' pregnancy intentions and to the combined intentions of both parents. Samples for analysis comprised 3,126 singleton children younger than 36 months, including 732 for whom paternal data were available.

Sixty-five percent of children were toddlers (aged 12–35 months) and the remainder were infants. They were roughly evenly distributed with respect to birth order, and most had been born 2–4 years after their next older sibling. The majority had been born to women who were married, were younger than 30 and had no more than a primary education. Substantial proportions had mothers who had not used contraceptives before the pregnancy (66%) or received maternal health services during pregnancy or delivery (22%). The majority (69%) lived in households with a piped water source, but 78% lacked a flush toilet. Fifty-nine percent lived in urban areas, and 41% in the high-elevation Altiplano region.

According to mothers' reports, 46% of children had been intended at conception, 33% had been unwanted, and 21% had been mis-timed. According to fathers' reports, these proportions were 23% and 16%, respectively.

Overall, 22% of children were stunted. Stunting was more common among children whose conception had been unwanted by their mothers than among those whose conception had been intended (29% vs. 19%). Those proportions were 37% vs. 18%, respectively, when fathers' reports were used. The prevalence of stunting was higher among toddlers (25–32%) than among infants (11%), among children born to contraceptive nonusers (26%) than among those born to users of modern methods (14%), and among children whose mothers had not used maternal health services both during pregnancy and at delivery (28–36%) than among those whose mothers had used maternal services at both points (14%). The prevalence of stunting rose with mother's age and child's birth order, and tended to decrease as mother's educational level and last pregnancy interval increased. It was higher among children living in poor conditions, rural areas and the Altiplano region (20–38%) than among children in more favorable environments (9–18%). These factors were all associated with childhood stunting at the bivariate level.

These patterns persisted in more detailed analyses stratified by mothers' pregnancy intentions. For all variables, stunting was more common among children born as the result of unwanted pregnancies than among those whose conception had been intended. Moreover, the link between stunting and unwanted pregnancy was stronger among toddlers than among infants (risk ratios, 1.7 and 1.5, respectively). The prevalence among children whose conception had been mistimed tended to fall between those for the other two groups; however, the relationship between mistimed status and stunting was rarely significant.

Multivariate analyses indicated that toddlers had a 30% higher risk of stunting if their mothers reported that their conception had been unwanted or mistimed rather than intended (risk ratio, 1.3 for each). In addition, the risk of stunting was significantly higher among toddlers whose conception had been reported as mistimed by both parents than among those whose conception had been reported as intended by both parents (3.8).

The researchers acknowledge the limitations of cross-sectional data and retrospective reports of pregnancy intentions. They suggest that the lack of a significant association [between mothers' intentions and stunting] in infants may be due to the protective effect of breastfeeding. The researchers note that improving policies and programs aimed at reducing unwanted and mis-timed pregnancies in Bolivia may decrease the prevalence of stunting in children. Moreover, they say, children born to parents reporting unintended pregnancies should be monitored for stunting and interventions should be in place to prevent stunting.—R. MacLean


1. Shapiro-Mendoza C et al., Parental pregnancy intention and early childhood stunting: findings from Bolivia, International Journal of Epidemiology, 2005, 34(2): 387– 396.