IJE Advance Access originally published online on November 23, 2004
International Journal of Epidemiology 2005 34(2):387-396; doi:10.1093/ije/dyh354
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Article |
Parental pregnancy intention and early childhood stunting: findings from Bolivia
1 School of Public Health, University of Texas Health Science Center at Brownsville, TX, USA
2 School of Public Health, University of Texas Health Science Center at Brownsville, TX, USA
* Corresponding author. Maternal and Infant Health Branch Division of Reproductive Health Center for Disease Control and Prevention Mailstop K-23; 4770 Buford Highway, NE Atlanta, GA 30341-3717, USA. E-mail: ayn9{at}cdc.gov
 |
Abstract |
|---|
 Top Abstract MethodsData analysisResultsDiscussionReferences |
|---|
Background This study examined the impact of maternally reported pregnancy intention, differentiating unwanted and mistimed pregnancies, on the prevalence of early childhood stunting. Additionally, it examined the influence of paternal pregnancy intention status.
Methods Data were collected from a nationally representative sample of women and men interviewed in the 1998 Bolivia Demographic and Health Survey. The sample was restricted to lastborn, singleton children younger than 36 months who had complete anthropometric information. Multivariable logistic regression examined the association between pregnancy intention and stunting.
Results Children from unwanted and mistimed pregnancies comprised 33% and 21% of the sample, respectively. Approximately 29% of the maternally unwanted children were stunted as compared to 19% among intended and 19% among mistimed children. Children 12–35 months (toddlers) from mistimed pregnancies (adjusted prevalence risk ratio [PRadj] 1.33, 95% confidence interval [CI]: 1.03–1.72) and unwanted pregnancies (PRadj 1.28, 95% CI: 1.04–1.56) were at about a 30% greater risk for stunting than children from intended pregnancies. Infants and toddlers with both parents reporting them as unwanted had an increased risk of being stunted as compared with children both of whose parents intended the pregnancy. No association was found for infants less than 12 months.
Conclusions Reducing unintended pregnancies in Bolivia may decrease the prevalence of childhood growth stunting. Children born to parents reporting mistimed or unwanted pregnancies should be monitored for growth stunting, and appropriate interventions should be developed. Measurement of paternal pregnancy intention status is valuable in pregnancy intention studies.
Keywords Bolivia, pregnancy, unwanted, family planning services, fertility, fathers, mothers, malnutrition
Accepted 14 September 2004
Despite international support for reproductive rights, Latin American mothers cannot effectively control their fertility and report more than one-third of their pregnancies as unwanted or mistimed.1,2 Parents who have unwanted and mistimed pregnancies may have fewer socioeconomic resources and lack social support, affecting children's susceptibility to poor nutrition, frequent illness, and infections. Additionally, a parent's feelings towards a pregnancy may adversely impact the child's health when these feelings contribute to conscious or unconscious neglect of the index child, resulting in inadequate nutrition, lack of parental bonding, and inattention to health care needs.3 These effects influence the process of becoming stunted, probably beginning in the perinatal period and continuing into childhood.
Being stunted, defined as decelerated or arrested linear growth, is a good indicator of long-term malnutrition. Other anthropometric indicators such as being wasted or being underweight have distinct aetiologies and are not good measurements of long-term morbidity and malnutrition.4–6 Stunted children are at a greater risk for impaired cognitive development and reduced neuromotor functioning, factors that contribute to poor academic performance and reduced work productivity in adulthood.5,7,8 Additionally, stunted or short-statured women have smaller pelvises placing them at a higher risk for obstetrical complications such as obstructed labor, which impinge upon maternal and infant survival.9–11
The global prevalence of stunting in early childhood ranges from 5 to 65% in developing countries.5 The prevalence of stunting in Latin America is 13%. In Bolivia, the prevalence of stunting has decreased from 34% in 1989 to 19% in 1998.12 Although this 44% decline is promising, Bolivia continues to have one of the highest prevalences of stunting in Latin America.5
Several US studies have found associations between unwanted pregnancies and adverse health or behaviours such as low birthweight, mortality, never breastfeeding, and poor prenatal care.13–23 However, there is little evidence for an association between mistimed pregnancies and adverse outcomes in these studies and in the data from poorer countries.13,14,24,25 High rates of mistimed and unwanted pregnancy in developing countries make it important to study its consequences on child growth.
Our analysis compared the prevalence of stunting among Bolivian children from birth to 36 months of age whose mothers retrospectively reported their pregnancy intention at conception. Since unwanted and mistimed pregnancies measure different dimensions of pregnancy intention, we assess them separately.26 Moreover, little is known about how paternal pregnancy intention affects child health, so we explored the joint effect of maternal and paternal pregnancy intention status on stunting in a sub-analysis of male partners.
|
Methods |
|---|
|
TopAbstractMethodsData analysisResultsDiscussionReferences |
|---|
We analysed data from the 1998 Bolivia Demographic and Health Survey (DHS), a nationally representative sample of women of childbearing age (15–49 years) and their male partners. The sampling design was a two-stage stratified multi-cluster sample.12 The men's sample was smaller and selected from the households of eligible women. The response rate was 95% for women and 89% for men.12
For this analysis, the dataset was restricted to 3447 surviving, lastborn singleton children who were younger than 36 months at the time of maternal interview. From this we excluded 153 children due to incomplete birth date information, 165 children with missing or improbable anthropometric measurements, and 3 children with missing responses for maternal pregnancy intention. The final sample size was 3126 children of whom 732 had information on the father's pregnancy intention status that was used for a sub-analysis. This sub-analysis was limited to married couples (including those living as married).
Data collection activities included household interviews using a pretested questionnaire and anthropometric measurements by trained field workers. The study imposed DHS-standardized quality control procedures, such as intensive training, adequate field supervision, and rigorous data entry procedures.1227–29
To measure pregnancy intention, the questionnaire asked respondents:
‘At the time you[your partner] became pregnant with [name of last born child], did you want to become pregnant then, did you want to wait until later, or did you want no more children at all?’‘Intended’ pregnancies were defined as those pregnancies to parent(s) who wanted the pregnancy, ‘unwanted’ referred to pregnancies to parent(s) not wanting more children, and ‘mistimed’ were those pregnancies that were wanted, but at a later time. The questions and definitions used here are standard for reproductive health surveys worldwide.13
A stunted child was defined as being more than two standard deviations (SD) below the reference median for height-for-age using the 1978 NCHS/WHO growth reference curves and corresponding Z-scores as the reference median, as per the recommendations of the World Health Organization.30
|
Data analysis |
|---|
|
TopAbstractMethodsData analysisResultsDiscussionReferences |
|---|
To compare the proportion of stunted children among unwanted and mistimed pregnancies to intended pregnancies, we used the prevalence risk ratio. We explored the role of selected covariates on stunting and growth using stratified analyses. We performed multivariable logistic regression analysis to estimate the prevalence risk ratio between pregnancy intention and stunting, stratifying by child age and adjusting for confounding. The main exposure variable, maternal pregnancy intention status, and socially and biologically important explanatory variables (birth order, maternal education, health service use, and latrine type) were included in the initial model. Using forward selection, other potential confounders were included if their addition to the model changed the estimated prevalence risk ratio by more than 10%.3132–34 This procedure allows for a more parsimonious public health model. This analysis produced estimates of the odds ratio that were transformed to prevalence risk ratios and 95% confidence intervals (CIs).35,36 For all analyses, we accounted for the cluster effects of the complex sampling design and applied sample weights to the strata using Stata's svy commands.37
Covariates considered for multivariable analysis were based on logical associations in the conceptual model, risk factors identified in the literature, and the results of stratified analyses. They included child's birth order, child's age, socioeconomic status (water access and type of latrine), maternal education level, preceding birth interval, use of health services (an index measure of use of prenatal care and place of delivery), marital status, age of mother, contraceptive use prior to pregnancy, urban vs rural residence, and geographic region (Altiplano, Valle, Llano). We assessed geographic region as a proxy for altitude because evidence regarding the association of altitude with stunting was unavailable for households.38–43
|
Results |
|---|
|
TopAbstractMethodsData analysisResultsDiscussionReferences |
|---|
Characteristics of the sample used for analysis are described in Table 1. The overall prevalence of stunting was 22.0%, of which 7.6% was severe stunting (below 3 SD from the reference median). According to the maternal report, the prevalence of mistimed pregnancies was 20.9%, compared with 32.7% unwanted pregnancies. The proportion of fathers reporting mistimed pregnancies was 16.1% compared with 22.8% reporting unwanted pregnancies.
|
Prevalence risk ratios for stunting by selected characteristics are also shown in Table 1. Nearly 40% of children were under 12 months of age. Mothers tended to be between 20 and 29 years of age, had not completed primary school, and were married when interviewed. Excluding firstborns (about one-quarter of the sample), most children were born between two and four years after their next older sibling. Most mothers had not used any contraception prior to the pregnancy, but over three-quarters reported using health services. In examining socioeconomic status indicators, 69.4% of children lived in households with piped water, but only 21.7% had flush toilets. Urban residence accounted for 59.1% of the study population and 41.0% resided in the Altiplano region.
The prevalence of stunting was highest for children aged 12 months and older, especially those 18–23 months old. The prevalence of stunting increased with increase in maternal age, but decreased with higher levels of maternal education. Stunting prevalence increased with birth order, short preceding birth intervals (less than 48 months and excluding firstborns), non-use of maternal health services or modern contraceptives, and for those living in poorer environmental conditions (having no flush toilet or piped water). In addition, the prevalence of stunting in children residing in rural areas was nearly twice that of children in urban areas, and was highest for children in the Altiplano region (the region having the highest altitude).
Infants (children under 12 months of age) had the lowest prevalence of stunting, regardless of intention status, while the prevalence in toddlers (children 12–35 months) ranged from 23.8% to 38.1%, with those from unwanted pregnancies having the highest prevalence in each age grouping (Table 2). Stratified analysis showed that for all maternal sociodemographic variables, the prevalence of stunting increased in the unfavorable sociodemographic categories (older maternal age, less education, higher birth order, lack of health service use, lack of sanitary latrine, rural or Altiplano residence) for all pregnancy intention groups. Among women with unwanted pregnancy, there is evidence that stunting is less prevalent in regular users of health care (19.5%) compared with non-users (38.4%). In all strata, compared with intended pregnancies, unwanted pregnancies resulted in an elevated prevalence risk ratio for stunting. In contrast, except for maternal age >40 years and having a flush toilet, mistimed pregnancies did not have an increased prevalence risk ratio of stunting in the stratified analysis.
|
There were no meaningful differences between stratum specific prevalence risk ratios for child age, maternal age, maternal education, birth order, health service use, rural vs urban residence, region, or type of toilet. In other words, there was no evidence of effect measure modification for these variables and this was confirmed with the likelihood ratio test for multiplicative interaction. However, because risk factors for stunting appear to differ between infants and toddlers, possibly due to the protective effect of breastfeeding, we report age-specific estimates in multivariate analysis. Additionally, most children may be in the process of stunting during infancy, but by definition do not become stunted until later.
Table 3 presents the risk of stunting associated with pregnancy intention as reported by the mother. When controlling for maternal education, birth order, type of toilet, and health service use, the effect of unwanted pregnancy on stunting was weakened. It was strengthened for mistimed pregnancy. In multivariable logistic regression, toddlers from mistimed pregnancies (adjusted prevalence risk ratio [PRadj] 1.33, 95% confidence interval [CI]: 1.03–1.72) and unwanted pregnancies (PRadj 1.28, 95% CI: 1.04–1.56) were each at about a 30% greater risk for stunting than children from intended pregnancies (Table 3). For infants, however, there was no difference in risk of stunting by pregnancy intention.
|
The risk of stunting associated with pregnancy intention as reported by the mother and father is shown in Table 4. Agreement between maternally and paternally reported pregnancy intentions was poor (weighted kappa, 0.28).44 Compared with children of the same age where both parents reported an intended pregnancy, having both parents report an unwanted pregnancy was associated with stunting for infants (PRadj 2.33, 95% CI: 0.96–5.68) and for toddlers (PRadj 1.64, 95% CI: 0.98–2.74) (Table 4). For infants, maternally reported mistimed and paternally reported unwanted pregnancies were associated with stunting (PRadj 5.33, 95% CI 0.94–30.14), but for toddlers this association was inconclusive due to imprecise measurement seen in the wide CIs. An association was found when both parents reported the pregnancy as mistimed for toddlers, but there were no stunted infants in this category. Other combinations of parental pregnancy intention status were inconclusive perhaps due to the small numbers in each stratum.
|
|
Discussion |
|---|
|
TopAbstractMethodsData analysisResultsDiscussionReferences |
|---|
Apart from the effects of socioeconomic status on stunting, maternally reported unwanted and mistimed pregnancy at conception was associated with stunting in toddlers, but not in infants. In Bolivia, the father's pregnancy intention status contributed to the association with stunting. For married couples, if pregnancies were unwanted by both the mother and father, infants and toddlers appeared to have an increased risk for stunting. Asking the father about his attitude regarding a pregnancy's intention is rare, but clearly of value in studies about pregnancy intention and subsequent health consequences.45,46
In stratified analysis, mistimed pregnancies tended to have similar prevalences of stunting as the intended pregnancies. When the prevalence was disparate, it often fell in the range between intended and unwanted pregnancies providing evidence that pregnancy intention occurs along a continuum. Because sociodemographic variables are directly associated with stunting and inversely related to mistimed pregnancy, for example, women with mistimed pregnancies tended to have factors protective against stunting such as being younger and having flush toilets, they act as negative confounders. This confounding accounts for the strengthened effect after adjustment.
Our study found that the child's age modified the association between maternal pregnancy intention and stunting, in that, an association was found for toddlers, but not for infants. This finding is not surprising since children often do not meet the conventional definition of below 2 SD from the international reference median for height-for-age until after two years of age.47 Using <1 SD as a cut-point for stunting, infants from unwanted pregnancies (PRadj 1.16, 95% CI: 1.00–1.35) had a 16% greater risk for stunting than intended pregnancies. No association was found with mistimed pregnancies.
The lack of a significant association in infants may be due to the protective effect of breastfeeding in reducing risk factors like diarrhoea and infectious diseases.4,41,47–52 Most of the sampled infants were ‘current breast feeders’ when interviewed (92% intended, 91% mistimed, and 96% unwanted, P = 0.17). In contrast, about 40% of toddlers were still breastfeeding (40% intended, 31% mistimed, and 44% unwanted, P = 0.23). Populations with different breastfeeding prevalences probably have different risk factors for stunting. Possibly older children, no longer benefiting from breastfeeding's protective effects, are more susceptible to infections and malnutrition exacerbated by neglect related to pregnancy intention.
An analysis of five DHS countries that included Bolivia, Marston, and Cleland14 reported an association between stunting and maternally reported mistimed or unwanted pregnancies only for Peru. Their analysis differed from the current analysis, in that, all children 0–60 months (not just lastborn and 0–35 months) were included. Our multivariate analysis stratified on child's age as opposed to adjusting for it. From our analysis, we believe that there is a window of risk, which their approach did not capture. Additionally, in our analysis women were only asked about events occurring in the three preceding years from the time their last child was conceived, thereby limiting the amount of recall bias due to remembering events that occurred greater than three years ago.
Montgomery et al.53 assessed the effects of maternal pregnancy intention on growth attainment in four developing countries. For the Dominican Republic, they found that compared with children aged 3–36 months of intended pregnancies, those who were unwanted had a deficit in height for age that was about a quarter standard deviation below the WHO/NCHS 1978 reference median [P < 0.05]. This statistically significant association was not found in the other countries studied (Egypt, Kenya, and Thailand) and results for mistimed pregnancies were not given. Our results appear to be consistent with the results from the Dominican Republic, another Latin American country with a moderate prevalence of stunting.
Cross-sectional information on stunting ascertained from currently living children might be unrepresentative of the growth status for the birth cohort. If selective survival occurs and more stunted children have died, it would thereby diminish any apparent magnitude of the effect. However, Boerma and others54, using longitudinal studies and DHS cross-sectional studies, found that survivor bias had little effect on estimates of stunting. In our study, the number of children who would have met the inclusion criteria but were excluded by death was 183 or 5.0%. The effect may also be diminished if unwanted or mistimed pregnancies were aborted. Unfortunately, reliable data for abortions in Bolivia is non-existent.
As in other studies that measure pregnancy intention retrospectively, post hoc rationalization is a concern. For example, a mother may report that her now living child was intended at the time of conception even though the pregnancy had been actually unwanted or mistimed. Most investigators agree that feelings or attitudes change from reports of ‘unwanted’ during pregnancy to ‘intended’ following birth, thereby reducing the effect of any association.1754–58 If a mother's responses were based on recalling the child's health at birth, the effect on the association could be reduced or magnified.59 Rewording the pregnancy intention question to have the mother refer to her feelings during conception helped minimize post hoc rationalization identified in DHS validity studies in Peru60 and the Dominican Republic.53
Presumably, normal birthweight children would take longer to become stunted than low birthweight infants. If birthweight was also related to pregnancy intention then it would be in the causal pathway between pregnancy intention status and being stunted. Contrary to this assumption, it has been reported that stunted children at two years usually have normal birthweights.61 Even so, it would have been interesting to examine stratum-specific effects of birthweight status in this study, but information was lacking on infant's birthweight (missing more than 35% of data on birthweight). As an alternative, we did a sub-analysis using those who reported hospital and clinic deliveries (fewer than 7% missing birthweight data). We found that compared with intended pregnancies, unwanted pregnancies (PRadj 1.58, 95% CI: 0.91–2.71) may have been more likely to be low birthweight (<2500 g). No association was found for mistimed pregnancies. This sub-analysis was not representative of the entire population and is reported tentatively.
Finally, we cannot rule out bias resulting from uncontrolled confounders that were not available in the data, such as depression. If pregnancy intention status was associated with postpartum depression, this would strengthen the effect of unwanted and mistimed pregnancies on stunting.
This study provides evidence that reducing unwanted and mistimed pregnancies in Bolivia may decrease the prevalence of stunting in children. It adds to the growing body of evidence that unwanted or mistimed pregnancies are detrimental to children13–23, especially when both father's and mother's intention status are considered. This study provides more evidence for policy makers planning interventions to better target women and men at risk for unintended pregnancy. In addition, children born to parents reporting unintended pregnancies should be monitored for stunting and interventions should be in place to prevent stunting. Finally, it suggests that child survival organizations should target parents who have unintended pregnancies or target services that prevent having unwanted and mistimed pregnancies, for example, providing effective and accessible family planning to women and men.
KEY MESSAGES
|
|
References |
|---|
|
TopAbstractMethodsData analysisResultsDiscussionReferences |
|---|
1 Bellamy C. The State of the World's Children, 2001. Oxford: Oxford University Press for UNICEF, 2001.
2 International Conference on Population and Development. Population and development: programme of action adopted at the International Conference on Population and Development, Cairo, September 5–13, 1994. New York: United Nations, Department for Economic and Social Information and Policy Analysis, 1995.
3 Parnell AM, DaVanzo J, Foege W, National Research Council (U.S.). Working Group on the Health Consequences of Contraceptive Use and Controlled Fertility, National Research Council (U.S.). Committee on Population, National Research Council (U.S.). Commission on Behavioral and Social Sciences and Education. Contraceptive use and controlled fertility: health issues for women and children: background papers. Washington, DC: National Academy Press, 1989.
4 World Health Organization. Physical status: the use and interpretation of anthropometry. Report of a WHO Expert Committee. World Health Organ Tech Rep Ser 1995;854:1–452.[Medline]
5 de Onis M, Frongillo EA, Blossner M. Is malnutrition declining? An analysis of changes in levels of child malnutrition since 1980. Bull World Health Organ 2000;78:1222–33.[Web of Science][Medline]
6 Shrimpton R, Victora CG, de Onis M, Lima RC, Blossner M, Clugston G. Worldwide timing of growth faltering: implications for nutritional interventions. Pediatrics 2001;107:E75.[CrossRef][Medline]
7 Mendez MA, Adair LS. Severity and timing of stunting in the first two years of life affect performance on cognitive tests in late childhood. J Nutr 1999;129:1555–62.
8 Pelletier DL, Frongillo EA Jr, Habicht JP. Epidemiologic evidence for a potentiating effect of malnutrition on child mortality. Am J Public Health 1993;83:1130–33.
9 Koblinsky MA. Beyond maternal mortality—magnitude, interrelationship, and consequences of women's health, pregnancy-related complications and nutritional status on pregnancy outcomes. Int J Gynaecol Obstet 1995;48(Suppl):S21–S32.
10 Camilleri AP. The obstetric significance of short stature. Eur J Obstet Gynecol Reprod Biol 1981;12:347–56.[CrossRef][Web of Science][Medline]
11 WHO Collaborative Study. Maternal anthropometry and pregnancy outcomes. Bull World Health Organ 1995;73(Suppl):1–98.[Web of Science][Medline]
12 Gutiérrez Sardán M, Ochoa LH, Gómez Vargas A. Bolivia encuesta nacional de demografía y salud, 1998. Calverton, MD: Macro International DHS Program, 1998.
13 Brown SS, Eisenberg L (eds). The Best Intentions: Unintended Pregnancy and the Well-Being of Children and Families. Washington, DC: National Academy Press, 1995.
14 Marston C, Cleland J. Do unintended pregnancies carried to term lead to adverse outcomes for mother and child? An assessment in five developing countries. Popul Stud (Camb) 2003;57:77–93.
15 Bustan MN, Coker AL. Maternal attitude toward pregnancy and the risk of neonatal death. Am J Public Health 1994;84:411–14.
16 Joyce TJ, Grossman M. Pregnancy wantedness and the early initiation of prenatal care. Demography 1990;27:1–17.[Web of Science][Medline]
17 Joyce T, Kaestner R, Korenman S. The stability of pregnancy intentions and pregnancy-related maternal behaviors. Matern Child Health J 2000;4:171–78.[CrossRef][Medline]
18 Joyce TJ, Kaestner R, Korenman S. The effect of pregnancy intention on child development. Demography 2000;37:83–94.[Web of Science][Medline]
19 Kost K, Landry DJ, Darroch JE. The effects of pregnancy planning status on birth outcomes and infant care. Fam Plann Perspect 1998;30:223–30.[CrossRef][Web of Science][Medline]
20 Baydar N. Consequences for children of their birth planning status. Fam Plann Perspect 1995;27:228–34, 245.[CrossRef][Web of Science][Medline]
21 Hummer RA, Schmertmann CP, Eberstein IW, Kelly S. Retrospective reports of pregnancy wantedness and birth outcomes in the United States. Soc Sci Q 1995;76:402–18.
22 Hellerstedt WL, Pirie PL, Lando HA et al. Differences in preconceptional and prenatal behaviors in women with intended and unintended pregnancies. Am J Public Health 1998;88:663–66.
23 Taylor JS, Cabral HJ. Are women with an unintended pregnancy less likely to breastfeed? J Fam Pract 2002;51:431–36.[Web of Science][Medline]
24 Chinebuah B, Perez-Escamilla R. Unplanned pregnancies are associated with less likelihood of prolonged breast-feeding among primiparous women in Ghana. J Nutr 2001;131:1247–49.
25 Eggleston E, Tsui AO, Kotelchuck M. Unintended pregnancy and low birthweight in Ecuador. Am J Public Health 2001;91:808–10.[Abstract]
26 Klerman LV. The intendedness of pregnancy: a concept in transition. Matern Child Health J 2000;4:155–62.[CrossRef][Medline]
27 Macro International Inc. Model ‘B’ questionnaire with commentary for low contraceptive prevalence countries. Columbia, MD: Macro International Inc., 1995.
28 Macro International Inc. Interviewer's manual for use with model ‘B’ questionnaire for low contraceptive prevalence countries. Columbia, MD: Macro International Inc., 1997.
29 Macro International Inc. Supervisor's and editor's manual for use with model ‘A’ and ‘B’ questionnaires. Columbia, MD: Macro International Inc., 1997.
30 Dibley MJ, Goldsby JB, Staehling NW, Trowbridge FL. Development of normalized curves for the international growth reference: historical and technical considerations. Am J Clin Nutr 1987;46:736–48.
31 Rothman KJ, Greenland S (eds). Modern Epidemiology. 2nd edn. Philadelphia: Lippincott-Raven, 1998.
32 Rothman KJ. Epidemiology: An Introduction. New York: Oxford University Press, 2002.
33 Kleinbaum DG, Kupper LL, Morgenstern H. Epidemiologic Research: Principles and Quantitative Methods. Belmont, CA: Lifetime Learning Publications, 1982.
34 Maldonado G, Greenland S. Simulation study of confounder-selection strategies. Am J Epidemiol 1993;138:923–36.
35 Zhang J, Yu KF. What's the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. JAMA 1998;280:1690–91.
36 Chronister K. Estimating relative risk using logistic regression with dichotomous and continuous covariates (Master's thesis). Houston: University of Texas Health Science Center, 2000.
37 StataCorp. Stata Statistical Software, Release 7.0. College Station, TX: Stata Corporation, 2001.
38 Haas JD, Moreno-Black G, Frongillo EA Jr et al. Altitude and infant growth in Bolivia: a longitudinal study. Am J Phys Anthropol 1982;59:251–62.[CrossRef][Web of Science][Medline]
39 Stinson S. The effect of high altitude on the growth of children of high socioeconomic status in Bolivia. Am J Phys Anthropol 1982;59:61–71.[CrossRef][Web of Science][Medline]
40 Yip R, Binkin NJ, Trowbridge FL. Altitude and childhood growth. J Pediatr 1988;113:486–89.[CrossRef][Web of Science][Medline]
41 Eveleth PB, Tanner JM. Worldwide Variation in Human Growth. 2nd edn. New York: Cambridge University Press, 1990.
42 Giussani DA, Phillips PS, Anstee S, Barker DJ. Effects of altitude versus economic status on birth weight and body shape at birth. Pediatr Res 2001;49:490–94.[Web of Science][Medline]
43 Harris NS, Crawford PB, Yangzom Y, Pinzo L, Gyaltsen P, Hudes M. Nutritional and health status of Tibetan children living at high altitudes. N Engl J Med 2001;344:341–47.
44 Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–74.[CrossRef][Web of Science][Medline]
45 Montgomery MR. Comments on men, women, and unintended pregnancy. Popul Dev Rev 1996;22(Suppl):100–06.[CrossRef]
46 Zabin LS, Huggins GR, Emerson MR, Cullins VE. Partner effects on a woman's intention to conceive: ‘not with this partner’. Fam Plann Perspect 2000;32:39–45.[CrossRef][Web of Science][Medline]
47 Waterlow JC. Linear Growth Retardation in Less Developed Countries. New York: Raven Press, 1988.
48 Seward JF, Serdula MK. Infant feeding and infant growth. Pediatrics 1984;74:728–62.
49 Martorell R, Yarbrough C, Lechtig A, Habicht JP, Klein RE. Diarrheal diseases and growth retardation in preschool Guatemalan children. Am J Phys Anthropol 1975;43:341–46.[CrossRef][Web of Science][Medline]
50 Martorell R, Habicht J-P. Growth in early childhood in developing countries. In: Falkner FT, Tanner JM (eds). Human Growth: A Comprehensive Treatise. 2nd edn. New York: Plenum Press, 1986, pp. 241–62.
51 Adair LS, Guilkey DK. Age-specific determinants of stunting in Filipino children. J Nutr 1997;127:314–20.
52 Eveleth PB. Population differences in growth: environmental and genetic factors. In: Falkner FT, Tanner JM (eds). Human Growth: A Comprehensive Treatise. 2nd edn. New York: Plenum Press, 1986, pp. 221–39.
53 Montgomery M, Lloyd C, Hewett P, Heuvelin P. The consequences of imperfect fertility control for children's survival, health, and schooling. Demographic and Health Surveys Analytical Reports. Calverton, MD: Macro International Inc., 1997. Report no. 7.
54 Boerma JT, Sommerfelt AE, Bicego GT. Child anthropometry in cross-sectional surveys in developing countries: an assessment of the survivor bias. Am J Epidemiol 1992;135:438–49.
55 Sable MR. Pregnancy intentions may not be a useful measure for research on maternal and child health outcomes. Fam Plann Perspect 1999;31:249–50.[CrossRef][Web of Science][Medline]
56 Westoff CF, Ryder NB. The predictive validity of reproductive intentions. Demography 1977;14:431–53.[Web of Science][Medline]
57 Bankole A, Westoff CF. The consistency and validity of reproductive attitudes: evidence from Morocco. J Biosoc Sci 1998;30:439–55.[CrossRef][Web of Science][Medline]
58 Kost K, Landry DJ, Darroch JE. Predicting maternal behaviors during pregnancy: does intention status matter? Fam Plann Perspect 1998;30:79–88.[CrossRef][Web of Science][Medline]
59 Rosenzweig MR, Wolpin KI. Maternal expectations and ex-post rationalizations: The usefulness of survey information on the wantedness of children. J Hum Resour 1993;28:205–29.
60 Goldman N, Moreno L, Westoff CF. Collection of survey data on contraception: an evaluation of an experiment in Peru. Stud Fam Plann 1989;20:147–57.[CrossRef][Web of Science][Medline]
61 Waterlow JC. Introduction. Causes and mechanisms of linear growth retardation (stunting). Eur J Clin Nutr 1994;48(Suppl. 1):S1–S4.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||