QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Villar, J. Articles by de Onis, M. Search for Related Content PubMed PubMed Citation Articles by Villar, J. Articles by de Onis, M.

---

Supplement: Nutrition as a Preventive Strategy against Adverse Pregnancy Outcomes

Nutritional Interventions during Pregnancy for the Prevention or Treatment of Maternal Morbidity and Preterm Delivery: An Overview of Randomized Controlled Trials ^{1 ,2}

José Villar^*,3, Mario Merialdi^*, A. Metin Gülmezoglu^*, Edgardo Abalos, Guillermo Carroli, Regina Kulier^** and Mercedes de Onis

^* UNDP/UNFPA/WHO/World Bank Special Programme of Research, Development and Research Training in Human Reproduction, WHO, CH–1211 Geneva 27, Switzerland, Centro Rosarino de Estudios Perinatales (CREP), WHO Collaborative Center in Maternal and Child Health, Rosario 2000, Argentina, ^** Geneva Foundation for Medical Education and Research, Geneva, Switzerland and Department of Nutrition, WHO, CH–1211 Geneva 27, Switzerland

³ To whom correspondence should be addressed. E-mail: villarj{at}who.int.

	ABSTRACT

TOP ABSTRACT LITERATURE CITED

 
This overview assesses the effectiveness of nutritional interventions to prevent or treat maternal morbidity, mortality and preterm delivery. Cochrane systematic reviews and other up-to-date systematic reviews and individual randomized controlled trials were sought. Searches were carried out up to July 2002. Iron and folate supplements reduce anemia and should be included in antenatal care programs. Calcium supplementation to women at high risk of hypertension during pregnancy or low calcium intake reduced the incidence of both preeclampsia and hypertension. Fish oil and vitamins E and C are promising for preventing preeclampsia and preterm delivery and need further testing. Vitamin A and ß-carotene reduced maternal mortality in a large trial; ongoing trials should provide further evaluation. No specific nutrient supplementation was identified for reducing preterm delivery. Nutritional advice, magnesium, fish oil and zinc supplementation appear promising and should be tested alone or together in methodologically sound randomized controlled trials. Anema in pregnancy can be prevented and treated effectively. Considering the multifactorial etiology of the other conditions evaluated, it is unlikely that any specific nutrient on its own, blanket interventions or magic bullets will prevent or treat preeclampsia, hemorrhage, obstructed labor, infections, preterm delivery or death during pregnancy. The few promising interventions for specific outcomes should be tested or reconsidered when results of ongoing trials become available. Until then, women and their families should receive support to improve their diets as a general health rule, which is a basic human right.

KEY WORDS: • nutritional interventions • maternal morbidity • preterm delivery • randomized controlled trials

Hypertensive disorders of pregnancy, hemorrhage, severe anemia, sepsis, obstructed labor and unsafe abortion and its complications are the main direct causes of maternal death. Nutritional deficiencies of both macro- and micronutrients are common in women of reproductive age in developing countries and epidemiological and biological evidence suggest that acute or chronic specific nutritional deficiencies can contribute to severe maternal morbidity. For example, an inverse relationship exists between calcium intake and the incidence of hypertensive disorders of pregnancy (1) and the prostacyclin-thromboxane imbalance reported in preeclampsia; various nutritional interventions have been suggested, including supplementation with naturally occurring fish oil (n–3 fatty acid) (2).

Severe anemia due to iron and folate deficiencies can increase the risk of maternal death from heart failure and augment the damage caused by antepartum or postpartum hemorrhage. Evidence implicating vitamin A deficiency in the pathogenesis of nutritional anemias is increasing (3– 6). Importantly, a randomized community intervention trial of maternal vitamin A supplementation in Nepalese women showed a reduction in overall maternal mortality (7).

Zinc is required in protein synthesis and nucleic acid metabolism (8, 9) and deficiency is common in developing countries (3, 9, 10). Both vitamin A and zinc deficiencies may contribute to perinatal sepsis by impairing the physiological response to infections.

Preterm birth (<37 completed weeks) complicates approximately 5–10% of all deliveries worldwide and is the major cause of perinatal mortality and long-term physical and neurological morbidity both in developing and industrialized countries (11). In most developing countries nutritional deficiencies and infections coexist, affecting women of childbearing age; an epidemiological association has been reported between lower genital tract and periodontal (12) infections and preterm birth (13, 14). Malaria, also prevalent in some tropical countries, is associated with severe anemia, preterm birth and impaired fetal growth (15).

Whether specific nutritional deficiencies in an otherwise healthy woman can cause preterm birth is not clear, although many nutritional factors, such as fish oil, magnesium and calcium, could theoretically affect the mechanisms controlling the initiation of labor.

We summarize here the results of systematic reviews of randomized trials of nutritional interventions during pregnancy. We ask the question Does the available evidence from randomized controlled trials demonstrate that nutritional interventions during pregnancy have any effect on reducing maternal morbidity, mortality or preterm delivery?

Conceptual framework

We propose a conceptual framework for interpreting the results from the systematic reviews and trials (Fig. 1).

View larger version (17K): [in this window] [in a new window]   FIGURE 1  Factors influencing the interpretation of results of clinical trials evaluating nutritional interventions during pregnancy.

Results from observational studies or uncontrolled evaluations are likely to be confounded by the effect of several covariates, a population selection bias. Nutritional deficiencies and infections are widely prevalent in most developing countries and inner-city populations of industrialized countries. Women residing in these locations are also at highest risk for pregnancy complications including preterm delivery. Epidemiological studies indicate that nutritional deficiencies and infections can play a role in the pathogenesis of preterm birth, making interventions to prevent or treat these problems attractive for primary or secondary prevention of preterm birth. However, recent large randomized controlled trials for the treatment of asymptomatic or bacteriologically positive vaginal infections to prevent preterm birth have failed to demonstrate such a protective effect (19– 21). Furthermore, selection bias of better-off populations is always present, making it likely that intervention groups will have better outcomes. This is the same selection bias observed in the epidemiological association between the number of antenatal care visits and better pregnancy outcomes that were not confirmed by large randomized trials (22, 23).

    Timing of the insult and different fetal organ growth patterns. The effect of a nutritional deficiency or of nutritional interventions is very likely to be related to the timing of occurrence during gestation. This is partially due to the differential fetal organ growth patterns (24) [e.g., as shown by ultrasonography and magnetic resonance imaging of brain, liver, colon, biparietal diameter and femur length (25– 28)]. Fetal liver, lung and brain volumes expressed as percentages of the total fetal volume vary with gestational age, indicating that they contribute differently at different gestational ages in relation to their growth rate patterns (27, 28). Overall, by 26–28 wk gestation fetal length is 70% of its average value at term whereas mean weight is only 32% of its average value at term (29). The ratio between head and abdominal circumference evaluated by ultrasound (which relates linear growth with organ growth) tends to decline progressively during pregnancy, reflecting the increased growth of the fetal abdomen compared with the head (30). Relating fetal growth trajectories and timing and balance of nutrients to newborn anthropometry still requires extensive epidemiological and pathological research (31), although large studies have shown that neonatal morbidity appears to be related to body proportionality (32– 34).

    Timing of nutrient deposition in the mother and the effect on fetal growth. Differential timing of nutrient deposition (weight gain during pregnancy) and its body location may also influence nutrient transfer to the fetus and therefore the effect of nutritional interventions during pregnancy. Patterns of fat and fat-free mass deposition differ in terms of location (e.g., thigh and subscapular versus triceps skinfolds) and timing (first and second trimester versus third trimester) (35). Birth weight is associated more with maternal changes in thigh skinfolds and early gestation fat gain than with other body sites or pregnancy timing (35– 37).

The importance of the location of fat deposition is not a new concept and it is not specific to pregnancy. Therefore, nutritional interventions during pregnancy, even if they achieve good compliance, are unlikely to affect pregnancy outcomes if they do not consider these parameters.

    Length and amount of nutritional supplementation. The concept of timing can be extended to length of supplementation and amount achieved because it is unrealistic that chronic undernutrition during two or three decades of life will be overcome, in terms of reproductive performance, with only a few months of extra nutrient intake. We observed, in an uncontrolled study, that energy supplementation had a biologically substantive effect (mean difference in birth weight: 301 g) when it was provided to women during two consecutive pregnancies and the interim lactation period but only a modest effect when women were supplemented only during the index gestation (39).

    Pharmacological effect versus nutritional effect. Nutrients are provided to populations as supplements to food either to increase the intake in those with a deficiency (to prevent or treat functional outcomes) or to obtain a pharmacological, perhaps nonnutritional effect, in individuals with an adequate intake of the nutrient. It is important to differentiate these two situations. As an example, our work on calcium supplementation during pregnancy is often discussed in the context of the results of a large, methodologically sound trial conducted by the U.S. National Institutes of Health (40). In the original formal description of the calcium-blood pressure hypothesis in 1980 (41), we specifically referred to the causal association between calcium deficit and hypertensive disorders of pregnancy and the causal role of calcium deficiency in the occurrence of hypertensive diseases of pregnancy. The importance of our observation was that increasing calcium intake in populations with a deficit may reduce the incidence of preeclampsia.

A large trial aimed at reducing the rate of pregnancy-induced hypertension and preeclampsia included 1167 women with a mean baseline calcium intake of 650 mg/d [approximately half of the recommended calcium allowance (42)] who were randomly assigned to receive 2000 mg/d of calcium or a placebo (43). Women took 86% of the supplement on average, increased their urinary excretion of calcium and had reduced risk of pregnancy-induced hypertension and preeclampsia [for preeclampsia, 95% confidence intervals (CI)⁴ of the odds ratios included the unity] (42).

In contrast, the National Institutes of Health trial studied the effect of calcium supplementation (2000 mg/d) in 4336 women without a calcium intake deficiency (mean baseline intake: 1130 mg/d) to achieve a pharmacological, preventive effect rather than to correct a nutritional deficit (40). No preventive effect of calcium supplementation on the incidence rates of preeclampsia and eclampsia was observed. This example demonstrates that supplementing deficient and nondeficient populations are two different concepts, which must be remembered when interpreting the results of clinical trials evaluating nutritional interventions.

This confusion between a nutritional and a pharmacological effect is not without risk. For example, the 1975 Harlem, New York, trial provided pregnant women without protein deficiency with a high-protein supplement (44). This pharmacological intervention had a negative effect on the rates of small for gestational age and neonatal death, and birth weight was on average 73 g lower in infants born to supplemented mothers.

A similar situation occurred in 2002 regarding the design of a large randomized trial to evaluate the suggested effect (45) of antioxidants (vitamins C and E) on the prevention of preeclampsia (James Roberts, Magee-Women's Research Institute, University of Pittsburgh, personal communication, 2002). Should such a trial be conducted among women without deficiency, expecting a pharmacological effect, or should it consider women with some evidence of vitamins C and E deficiency?

    Intervention-specific outcomes versus overall outcomes for morbidity, mortality and birth weight. Identifying the primary outcome or the most specific outcome in reference to the nutrient being evaluated is also a key element. Maternal underweight may not be associated with pregnancy complications such as preeclampsia, gestational diabetes, placenta previa or abruptio placenta, yet underweight mothers have been shown to be at higher risk for having newborns with birth weight below the 5^th percentile and preterm delivery (46).

Magnesium sulfate is an effective intervention when used parentally for the treatment of preeclampsia (47) and eclampsia (48) but it did not have effect an on fetal growth and low birth weight because these outcomes were already present when treatment was initiated (47). To complicate the issue further, some treatments could improve clinical outcomes during pregnancy but adversely affect other nutritional outcomes such as birth weight and fetal growth (i.e., the use antihypertensive medications to treat preeclampsia reduces blood pressure but also birth weight) (49, 50).

Traditional substantive outcomes such as mortality, severe morbidity and newborn anthropometric measures were the primary outcomes considered in the trials reviewed. However, other more specific outcomes may prove as important as the traditional ones in addressing public heath issues. For example, a trial of maternal zinc supplementation during pregnancy conducted in Peru found no effect on birth weight of supplementing zinc-deficient pregnant women with 25 mg/d of zinc (51). However, maternal zinc supplementation had a positive effect on fetal femur diaphysis growth as assessed by ultrasonography and on the development of heart rate patterns during fetal life (51). The biological significance of these effects still needs to be determined.

Finally, we should address the question of the relative importance of the long-term outcomes of nutritional interventions compared with perinatal interventions. The extensive literature on the relationship between fetal nutrition and adult disease (52) and the limitations of the relationship (53) are still being debated, although evidence from randomized controlled trials is very limited (54, 55). A follow-up study of the effects of magnesium sulfate on the newborn is now underway.

    Heterogeneity of outcomes. In evaluating the effect of nutritional intervention on pregnancy outcomes, it is important to emphasize again the heterogeneous nature of the outcomes considered. For example, not all low birth weight infants are the same (preterm vs. intrauterine growth restriction) (56, 57) and not all preterm births result from the same cause (58); the biological markers examined so far provide evidence of the multifactorial etiology of preterm birth. Likewise, mortality, either maternal or perinatal, might be too crude and comprehensive an outcome to show any variation in relation of the specific effect of a nutritional intervention in pregnancy. Therefore, it is unlikely that a blanket nutritional intervention provided to women, some of whom may not even have a nutritional deficiency, will have a major effect overall in preventing preterm delivery, preeclampsia or mortality. Preeclampsia and preterm delivery should be subclassified under different etiological forms to enable more specific interventions to be identified (59).

Methodological issues of systematic reviews

A systematic review is a systematic search and critical evaluation of all primary studies answering the same question. Statistical methods may or may not be used to analyze and summarize the results of the trials included in the review. Meta-analysis involves the use of statistical techniques within a systematic review to integrate the results of the included studies. Systematic reviews are important tools that help clinicians, health providers, researchers and policymakers to summarize the existing information in order to make evidence-based decisions. Systematic reviews are different from narrative reviews because they are based on protocols, hypothesis and selection criteria prepared before conducting the review and examining the data.

We present here results based only on controlled trials with random allocation of eligible subjects to receive or not to receive one or more nutritional interventions or placebo. The results are assessed by comparing outcomes in the treatment and control groups. Because of random allocation, baseline characteristics for known and unknown possible confounding variables are expected to be distributed similarly between groups (18).

The ratio of risk in the intervention (exposed) group to the risk in the control (unexposed) group [relative risk (RR)] is the parameter used. RR = 1 indicates no difference between comparison groups. For harmful or undesirable outcomes, RR < 1 indicates that the intervention was effective in reducing the risk of that outcome. Meta-analysis can estimate a pooled RR for all the trials included in a systematic review. The CI is the range within which the true RR value is expected to lie with 95% degree of certainty. CIs are included also in the meta-analysis.

Although there is a tendency in the literature to force the results of a systematic review into a single pooled estimator using meta-analytical techniques, when substantial differences occur among trial results or in the face of statistical heterogeneity, a single estimate may be misleading. We present the results of the meta-analyses using only fixed-effect models. Fixed-effect models assume that there is a common effect and a random component (sampling error) that is responsible for differences among trial results (i.e., it assumes homogeneity of intervention effect). This approach provides inferences only about the set of trials under review, giving weight to each trial based on the within-study sampling variance. The individual study sample size and the number of events are the leading factors in the weight assigned to each trial in the pooled estimate of the relative risk. Larger trials are therefore given more weight.

In case of statistical heterogeneity (p < 0.10) (60) or if relevant clinical heterogeneity in trials' results was present, we did not force the results into a single summary estimation and present them stratified by a relevant prognosis variable or by individual trial (61). Funnel plots for evaluation of publication bias were produced for meta-analyses that included three or more clinical trials (62).

    Selection criteria. Systematic reviews of randomized controlled trials or individual trials of nutritional interventions during pregnancy were included. Interventions to prevent or treat serious maternal morbidity that were aimed at stopping labor or prolong pregnancy after a diagnosis of preterm labor were not included. Systematic reviews of interventions to prevent preterm delivery but were not exclusively nutritional (such as alternative antenatal care packages or social support during pregnancy) were also excluded.

    Type of participants. Participants in the trials were pregnant women with or without any prior risk at assessment or screening.

    Type of interventions and outcome measures. Any systematic review of randomized controlled trials or individual trials of a nutritional intervention during pregnancy was eligible if the primary objective was to prevent maternal mortality, morbidity or preterm delivery. Main outcome measures were maternal mortality, preeclampsia or eclampsia, anemia, hemorrhage and obstructed labor and delivery rate before 37 completed weeks.

    Search strategy. The sources of systematic reviews were the Cochrane Database of Systematic Reviews, Cochrane Controlled Trials Register and Database of Abstracts of Reviews of Effectiveness included in the 2002 issue 2 of the Cochrane Library. A literature search for reports of interventions not yet included in the reviews was also conducted. All included reviews were published either as full reviews or as protocols in the Cochrane Library. We did not make any addition to the meta-analysis of these Cochrane systematic reviews if a new trial was identified after the publication date of the systematic review but the results of any new trial were included in the text. The trials included in the systematic reviews reported on this overview and other more recent intervention trials are included in Annex Table 1 in the article by Villar et al. in this issue (63).

    Data extraction. Data extraction was done by one of the authors assigned to a specific topic and checked in a random subgroup by a second author. Settings in which trials were conducted, specific interventions tested and methods of allocation were also extracted and are available on request. Trials not included in the reviews were assessed individually for methodological quality according to standard criteria used in Cochrane reviews (64).

Results

    Nutritional interventions for the prevention of high blood pressure or preeclampsia. Controversies exist about the definition of hypertensive disorders during pregnancy, and several classifications have been suggested. Recently, the U.S. National High Blood Pressure Education Program Working Group on High Blood Pressure in Pregnancy (65) classified the hypertensive disorders during pregnancy as chronic hypertension defined as hypertension observable before pregnancy or diagnosed before 20 wk gestation; preeclampsia, which is a pregnancy-specific syndrome occurring usually after 20 wk gestation, determined by hypertension with proteinuria; preeclampsia superimposed on chronic hypertension; and pregnancy-induced hypertension or gestational hypertension, which is transient hypertension detected for the first time after midpregnancy if preeclampsia is not present at delivery and blood pressure returns to normal by 12 wk postpartum (a retrospective diagnosis). The system suggested by the International Society for the Study of Hypertension in Pregnancy defines hypertension as a diastolic blood pressure of 90 mmHg or above on two consecutive occasions at least 4 h apart or a single diastolic blood pressure of 110 mmHg or above. The definition of preeclampsia has the same criteria for high blood pressure but with the addition of significant proteinuria, usually at least 300 mg/24 h or 1+ on dipsticks (66). These two definitions and in some instances others were used in the trials reviewed.

Considerable evidence from observational and experimental studies links calcium intake and hypertension during pregnancy. However, no satisfactory explanation exists for the mechanisms involved in the calcium-mediated effect on blood pressure reduction. Parathyroid hormone was postulated to be involved in this relationship (67, 68). Alterations in extracellular calcium homeostasis in preeclampsia, such as hypocalciuria and decreased serum levels of calcitrol, have been reported (69, 70).

An additional role for nutrition in the genesis of preeclampsia could be the deficiency of antioxidant intake (71), specifically vitamins C and E. Vitamin C is central for the neutralization of both water-soluble and lipid-soluble free radicals in the aqueous compartment and must come from the diet. Vitamin E, a potent antioxidant, was suggested to play a role in preventing preeclampsia. Vitamin E serum levels are not usually reduced by dietary deficiencies and reductions are more likely due to consumption resulting from its antioxidant activity (72, 73).

Other nutritional factors can also contribute to oxidative stress. Hyperhomocysteinemia, a risk factor for preeclampsia and atherosclerosis (74), is related to altered endothelial function at least partly by the genesis of oxidative stress. Vitamins B-6 and B-12 and folic acid are involved at different steps in the metabolic pathway for removing or recycling homocysteine to methionine (75); dietary deficiencies of any of these micronutrients can increase circulating homocysteine (76, 77).

Preeclampsia is characterized by increased triglycerides that favor the formation of LDL. LDL has increased access to the subendothelial space, where it is sequestered from blood-borne antioxidants. The relevant role of triglycerides in the genesis of preeclampsia is indicated by their being increased long before the manifestation of the disease. Similarly, free fatty acids are increased in preeclampsia and this increment was observed months before the diagnosis (71, 78). Recent studies indicate that this effect may be secondary to altered copper binding by albumin to which large amounts of free fatty acids are bound. Unbound copper is a potent stimulator of free radical formation. Ordinarily this effect of copper is prevented by protein binding (quantitatively, primarily to albumin), but albumin binds copper differently when fatty acids are present and copper maintains its ability to participate in redox reactions (79, 80). Thus, increased free fatty acids appear to contribute to oxidative stress. Most of these nutritional mechanisms may be modified by diet, raising the possibility of nutritional prophylaxis of preeclampsia.

The ability to prevent hypertensive disorders of pregnancy is limited by lack of knowledge of its underlying etiology, including the possibility that preeclampsia and hypertension during pregnancy may be two different entities. Prevention is traditionally focused on identifying women at higher risk, followed by close clinical and laboratory monitoring to recognize the clinical symptoms of the disease in its early stages. These women and their pregnancies can then be selected for more intensive monitoring or delivery. Although these measures do not prevent disease, they may help to prevent some severe maternal complications and fetal sequelae. Nutritional interventions may be added to this strategy; evidence for such interventions is presented in the following sections.

Nutritional advice in pregnancy. We assessed the effects of advising pregnant women to increase their energy and protein intakes on maternal and fetal-infant morbidity and mortality. Nutritional advice appears to be effective in increasing pregnant women's energy and protein intake, but the implications for fetal, infant and maternal health cannot be judged from the available evidence. Among the trials included in the Cochrane review (81), preeclampsia prevention was assessed only in one small trial [See Annex Table 1 (63) Trial 1] that showed no beneficial effects (Table 1).

Protein and energy restriction for obese women. Epidemiological studies suggest that high maternal weight is positively associated with risk for preeclampsia (94– 96). Excessive weight gain during pregnancy has long been mentioned as a clinical sign of edema and impending preeclampsia. Protein-energy restriction for high weight for height or high weight gain during pregnancy (several definitions were used) was assessed in a Cochrane systematic review. Preeclampsia was evaluated in two trials [See Annex Table 1 (63) Trials 20,21] that showed no reduction in the risk of occurrence (84). The same negative pattern was seen for pregnancy-induced hypertension in three trials [See Annex Table 1 (63) Trials 19,20,21] (Table 1). This limited evidence suggests that protein-energy restriction for pregnant women who are overweight or exhibit high weight gain during pregnancy is unlikely to be beneficial and may be harmful to the developing fetus [i.e., result in intrauterine growth retardation (97)]. It is therefore surprising that clinicians still frequently recommend that pregnant women restrict their food intake in an attempt to prevent preeclampsia.

Salt restriction. From early pregnancy marked hemodynamic changes including a fall in vascular resistance and blood pressure and a rise in cardiac output are observed. To compensate for the increased intravascular capacity, the kidney retains more sodium and water (98). Apparently, the set point of sodium homeostasis shifts higher at the expense of an expansion of extracellular volume. In nonpregnant individuals, a strong positive association of sodium intake with blood pressure has been established, but the relationship between sodium intake and blood pressure in human pregnancy remains obscure.

For decades a low-salt diet has often been recommended as a treatment for edema, with the idea that restricting salt intake would treat and prevent preeclampsia. Later, not only was this practice questioned, but an even higher sodium intake was proposed for preeclampsia treatment and prevention (99, 100).

The concerns about the effect of a low-sodium diet during pregnancy on maternal nutritional status led researchers to investigate whether such changes could alter other nutrient intakes. The reduction in sodium intake also caused a significant reduction in the intake of energy, protein, carbohydrates, fat, calcium, zinc, magnesium, iron and cholesterol (101). Even if most women are no longer advised to alter their salt intake during pregnancy, this remains a practice in some clinical settings around the world.

A Cochrane review evaluated the effect of the advice about low dietary salt intake during pregnancy (85). This review included two trials [See Annex Table 1 (63) Trials 22,23] with data reported for women without preeclampsia, comparing nutritional advice to restrict dietary salt with advice to continue a normal diet. The review provides no information about the effects of advice of restricted salt intake for the treatment of preeclampsia. No effect was found in preventing preeclampsia or pregnancy-induced hypertension (Table 1). Women's dietary preferences were not reported, but the authors presumed that a low-salt diet was not very palatable and therefore difficult to follow (85).

Calcium. A role for altered calcium metabolism in the pathogenesis of preeclampsia has been suggested based on epidemiological evidence linking low dietary levels of calcium with increased incidence of the disease (1). Supported by these observations, several modifications in calcium metabolism were observed in preeclamptic women and in calcium-supplemented mothers (67, 68, 102).

For a Cochrane review of calcium supplementation during pregnancy, the authors prespecified stratified analysis taking into account women's risk of hypertensive disorders of pregnancy (low vs. increased) and baseline dietary calcium intake (low: <900 mg/d vs. adequate: 900 mg/d) (86). High blood pressure with or without proteinuria was evaluated in 10 trials involving 6634 women. The rate of high blood pressure was lower with calcium supplementation (typical RR 0.81; 95% CI: 0.74–0.89) but there was heterogeneity in the magnitude of the effects across the subgroups. The effect was considerably greater in women at high risk of developing hypertension than in those at low risk. The effect was also greater in those with low baseline dietary calcium intake than in those with adequate calcium intake (Table 1).

Overall, the risk of preeclampsia, evaluated in 11 trials [See Annex Table 1 (63) Trials 24–34] (typical RR 0.70; 95% CI: 0.58–0.83), was reduced (86). When predefined subgroups are considered, a significant reduction was noted in women with low baseline dietary calcium intake but not in those with adequate calcium intake (Table 1). Preeclampsia was considerably reduced in women at high risk of hypertension and less consistently in those at low risk of hypertension (Table 1).

The results from the largest trial conducted by the U.S. National Institutes of Health, which studied only low-risk women with adequate calcium in their baseline diet, all of whom received low-dose calcium supplementation as part of their routine antenatal care, showed no significant effect on hypertension or preeclampsia [See Annex Table 1 (63) Trial 26]. Based on these results, populations with adequate dietary calcium intake are presently not encouraged to take routine calcium supplementation during pregnancy. However, evidence from this review supports the view that calcium supplementation might benefit women at high risk or women with low dietary calcium intake.

A large (8500 women), double-blind randomized controlled trial is now being conducted by the World Health Organization (WHO) and seven research institutions worldwide, supplementing women with low calcium intake (<600 mg/d) with 1500 mg/d of calcium after the week 20 of pregnancy. The results are expected by the end of 20za03.

Iron and folate. Numerous trials involving various populations of pregnant women with different hemoglobin levels evaluated the effects of supplementation with iron, folate or both on several outcomes; some trials included hypertensive disorders of pregnancy. A Cochrane review of 2 trials [See Annex Table 1 (63) Trials 35,36] involving women with normal hemoglobin levels, in which iron and folic acid were compared with no supplement, showed no effect on the occurrence of gestational hypertension (preeclampsia was not evaluated) (Table 1) (89). Another Cochrane review that included two trials [See Annex Table 1 (63) Trials 52,53] involving pregnant women already receiving iron, in which women were enrolled to receive folic acid or no treatment/placebo, also showed no effect on preventing gestational hypertension (Table 1) (90). Therefore, evidence shows that iron and folate supplementation is not effective in preventing hypertensive disorders during pregnancy, although they may be prescribed for other established beneficial effects of pregnancy such as prevention and treatment of anemia (90, 92).

Magnesium. Magnesium is one of the essential minerals needed by humans in relatively large amounts. Magnesium works with many enzymes regulating temperature and protein synthesis as well as maintaining electrical potentials in nerves and muscle membranes (103). Magnesium occurs widely in many foods; dairy products, breads and cereals, vegetables and meats are all good sources. Not surprisingly, frank clinical magnesium deficiency has never been reported to occur in healthy individuals who eat standard diets.

Dietary surveys during pregnancy consistently demonstrate that many women, especially those from disadvantaged backgrounds, have intakes of magnesium below recommended levels (104). Observational studies based on medical records reported that magnesium supplementation during pregnancy was associated with a reduced risk of fetal growth retardation and preeclampsia (105) and that magnesium intake was positively associated with birth weight (106). Results from this and other epidemiological studies encouraged researchers to conduct randomized trials to evaluate the potential benefits of magnesium supplementation during pregnancy.

A Cochrane review including two trials [See Annex Table 1 (63) Trials 64,67] showed no apparent effect of magnesium supplementation on the prevention of preeclampsia (mean supplement dose 365 mg and 500 mg) (Table 1) (87). These results may have been confounded by the fact that in the largest trial [See Annex Table 1 (63) Trial 67], all women received a multivitamin and mineral preparation containing low doses of magnesium (mean supplement dose 100 mg). The methodological quality of these trials was poor, especially related to concealment of treatment allocation. The authors of the review concluded that dietary magnesium supplementation of pregnant women cannot be recommended for routine clinical practice because of the poor methodological quality of the current evidence (87). This is, of course, unrelated to the effectiveness of parenteral magnesium sulfate for treatment of preeclampsia (47) and eclampsia (48) demonstrated in two large randomized controlled trials.

Fish oil. An early meta-analysis of controlled clinical trials of the effect of fish oil, rich in long chain n–3 fatty acids, on blood pressure in untreated nonpregnant subjects demonstrated a significant reduction in systolic and diastolic blood pressure but found no significant effect in normotensive subjects (107). Fish oil has been shown to interfere with prostaglandin metabolism, and its effect on blood pressure has often been assumed to be due to such interference. Epidemiological studies suggested that marine diets could have a preventive effect on early delivery and hypertensive disorders of pregnancy (108, 109).

Fish oil supplementation during pregnancy was evaluated in 2 trials [See Annex Table 1 (63) Trials 71,72] in an early Cochrane review that included 3 trials and showed no effect on pregnancy-induced hypertension (88). There was a statistically significant modest reduction in the rate of preeclampsia (Table 1), but this reduction was strongly influenced by a large nonrandomized trial conducted in 1942 in which vitamins and minerals were given to women in addition to fish oil [See Annex Table 1 (63) Trial 72].

Another four trials of fish oil supplementation involving more than 2000 women were published in the past few years (Table 2) and none were was included in the last Cochrane review. None of the four trials demonstrated any differences between groups in the incidence of hypertension during pregnancy or preeclampsia. The first trial, a randomized double-blind trial conducted in Leeds, England, [See Annex Table 1 (63) Trial 71], compared fish oil with placebo and involved women at high risk of pregnancy hypertension or intrauterine growth restriction (IUGR). These were defined by abnormal uterine Doppler at 24 wk gestation for primigravidas or history of pregnancy-induced hypertension, IUGR or stillbirths in previous pregnancies for multipara. No differences were noted in the incidence of both preeclampsia and pregnancy-induced hypertension.

The three-arm study conducted in Denmark compared fish oil with olive oil or no oil supplementation in healthy primiparous women, starting supplementation before 30 wk gestation [See Annex Table 1 (63) Trial 73]. No sound conclusion could be drawn about preeclampsia prevention because there was only one case in the trial, as reflected by the wide confidence intervals, or pregnancy hypertension when fish oil was compared with no treatment.

The fourth trial was a multicenter randomized clinical intervention program consisting of a series of four prophylactic and two therapeutic trials; the program was conducted in 19 centers in Europe [See Annex Table 1 (63) Trial 70a] and compared fish oil with olive oil. The prophylactic trials evaluated women with uncomplicated pregnancies and a history of preterm delivery (EARL-PD trial), IUGR (EARL-IUGR trial) and pregnancy-induced hypertension (EARL-PIH trial) and women with uncomplicated current twin pregnancies (TWINS trial). Pregnancy-induced hypertension was evaluated in the EARL-PIH trial and in the TWINS trial. For preeclampsia, no statistically significant differences were found neither the EARL-PIH trial or the TWINS trial. On the basis of current evidence, fish oil supplementation is not recommended during pregnancy for preeclampsia or pregnancy-induced hypertension.

Zinc. Zinc is critical to the expression of genetic potential (110) and has an important role in nucleic acid metabolism and protein synthesis (111). Thus, zinc deficiency is more likely to produce negative effects in periods of increased gene expression and rapid tissue growth, such as pregnancy, fetal life (112) and childhood, as suggested by the results of a meta-analysis of zinc supplementation trials in children showing an effect of zinc on growth (113). Maternal zinc deficiency is increasingly suggested as a possible public health problem (114). It has been associated in some epidemiological studies with reduced fetal growth and neurobehavioral development, preterm delivery, labor complications and delivery complications and was suggested to affect other postnatal outcomes such as immunological development, vitamin A status and postnatal growth (114).

The role of routine zinc supplementation during pregnancy was assessed in a Cochrane systematic review in 1997 (91). Since 1997, 3 new and methodologically sound clinical trials of maternal zinc supplementation conducted in Peru, Bangladesh and Nepal [See Annex Table 1 (63) Trials 74,75,82] have been published. One trial examined only the effect of maternal zinc supplementation in augmenting the effect of vitamin A in restoring night vision in vitamin A–deficient pregnant women [See Annex Table 1 (63) Trial 75].

Six other randomized trials of maternal zinc supplementation were conducted recently in Ecuador, Chile, Indonesia (West Java and East Java), Timor and Peru. None of these six studies has been published in journal yet, but results from two trials (West Java and Peru) were published as a doctoral thesis [See Annex Table 1 (63) Trials 76,81]. Therefore we include in the tables only the results of the two doctoral dissertations and discuss the findings of the other trials.

Of the seven studies included in the Cochrane review, four had pregnancy hypertension as an outcome [See Annex Table 1 (63) Trials 78,79,80,85]. Pregnancy-induced hypertension or preeclampsia was not found to be affected by zinc supplementation (Table 1) (91). The three studies published after the update of the review and the six unpublished studies do not present data on these outcomes. This was expected because no biological link between zinc and hypertension in adults is known.

Vitamins E and C. An oxidant-antioxidant imbalance has been suggested as a pathogenic factor involved in preeclampsia (115). Because vitamin E is one of the most important antioxidants in the human body components, its levels and their relation with circulating levels of lipids peroxides on preeclamptic women were intensively studied in recent years. As with other antioxidants, some studies found lower vitamin E levels in serum from women with gestational hypertension or preeclampsia than in serum in control women (116, 117). Increased ascorbate radical formation and ascorbate depletion were found in plasma from women with preeclampsia (118), but these findings were not supported by another study (119).

Antioxidants have been proposed as a potentially advantageous prophylactic measure for preeclampsia (72). One explanatory trial of combined antioxidant treatment (vitamin C, vitamin E and allopurinol) in women with established severe early-onset preeclampsia showed no substantial clinical benefit (120). However, there was a statistically non significant tendency toward prolongation of pregnancy in the antioxidant group (for delivery within the next 14 d: RR 0.68; 95% CI: 0.45–1.04).

The first randomized controlled trial on prophylactic vitamins E and C supplementation involved women at high risk of developing preeclampsia in London [See Annex Table 1 (63) Trial 87]. Women at increased risk (defined by authors as those having abnormal Doppler waveform in either uterine artery at 18–22 wk gestation or history of preeclampsia in the previous pregnancy) were randomly assigned to receive vitamin E and vitamin C supplements or a placebo at 16–22 wk gestation. The authors found a large significant reduction in the risk of developing preeclampsia in the supplemented group compared with the control group (Table 1). Although these findings are promising, they were from one small trial. The postulated role of vitamins C and E on preeclampsia is currently being evaluated in a large multicenter double-blind randomized trial in North America (James Roberts, Magee-Women's Research Institute, University of Pittsburgh, personal communication, 2002).

Vitamin A. The role of vitamin A in pregnancy induced-hypertension and preeclampsia is also a subject of controversy (121). Some clinical studies found significantly lower serum vitamin A levels in preeclamptic women than in healthy women in the third trimester (117). No trials have been published that assess the effect of vitamin A supplementation on pregnancy-induced hypertension or preeclampsia.

A double-blind, cluster-randomized trial of low-dose vitamin A or ß-carotene supplementation carried out in Nepal showed a 40% reduction in overall maternal mortality among vitamin A–supplemented women [See Annex Table 1 (63) Trial 92]. When differences in cause of deaths including preeclampsia and eclampsia were evaluated, the rates of these conditions were not different between supplemented and placebo groups.

Multiple micronutrients. The only trial to have supplemented pregnant women using multiple micronutrients was carried out in Mexico in a nutritionally deficient rural population (Teresa Gonzales-Cossio, Instituto Nacional de Perinalogía, personal communication, 2001). The primary outcome measures in this trial are newborn anthropometry, gestational age, early postnatal growth and micronutrient status of mothers and infants. Substantial information on preeclampsia and eclampsia is unlikely to come from this trial.

    Nutritional interventions for the treatment of hypertensive disorders of pregnancy and hypertension. Mild and uncomplicated chronic hypertension during pregnancy has a better prognosis than preeclampsia. However, risk for superimposed preeclampsia and possible complications is increased if preexisting renal disease or systemic illness is present (122, 123). The primary aim of therapy is to prevent hypertension-related complications and to avoid the progression to superimposed preeclampsia with worse prognosis. Overall, nonpharmacological management of mild hypertension during pregnancy remains controversial.

In a published review of interventions for management of mild chronic hypertension during pregnancy, no trials were found that compared nonpharmacological interventions with either pharmacological agents or no intervention in pregnant women (124). This comprehensive review identified 50 randomized controlled trials, but they involved either normotensive women or women with a history of preeclampsia. For the management of preexisting chronic hypertension during pregnancy, no relevant evidence could be located to assess the effects of nonpharmacological interventions, such as limiting activity, diet modifications or stress reduction.

In general, weight reduction during pregnancy, even in obese women, is not recommended to improve pregnancy outcomes (97). Even though obesity may be a risk factor for superimposed preeclampsia, there is no evidence that limiting weight gain during pregnancy reduces its occurrence (84).

Pregnant women with hypertension have lower plasma volume than do normotensive women, and the review of some studies suggests that the severity of hypertension correlates with the degree of plasma volume concentration (125). For this reason, sodium restriction is generally not recommended during pregnancy for the reduction of blood pressure. In addition, an increase in plasma volume concentration is a risk factor for IUGR. If, however, a pregnant woman with chronic hypertension is known to have salt-sensitive hypertension and has been treated successfully with a low-salt diet before pregnancy, it is reasonable to continue some sodium restriction for blood pressure control during pregnancy but not for preventing superimposed preeclampsia (65, 97). The effect of such approach on fetal and neonatal outcomes is still unknown.

In summary, no reliable information from well-designed randomized controlled trials assesses the best dietary approach for the management of preexisting hypertension during pregnancy.

    Nutritional interventions for the prevention of anemia- and hemorrhage-related outcomes. The systematic reviews of nutritional advice, protein-energy supplementation and energy restriction did not include any anemia- and hemorrhage-related outcomes.

Calcium. Calcium may inhibit both heme and nonheme iron absorption (126– 128). Dietary strategies are recommended to reduce the possible negative effect of calcium intake, such as taking calcium during meals with low iron content. Whether reduced iron absorption is of clinical significance during a finite period such as pregnancy and under pragmatic circumstances is not clear. We explored the incidence of two different cutoff points for hemoglobin values by week of gestation in women enrolled in the first large randomized controlled trial of calcium supplementation [See Annex Table 1 (63) Trial 24]. In this population, although iron supplementation was universally recommended, overall 35% of women in each group took iron supplements during their pregnancy. The incidence of hemoglobin <100 g/L or <110 g/L during pregnancy was very similar in the calcium and placebo groups. Hemoglobin values were also compared between calcium and placebo groups only for women who took iron supplements (post randomization analysis), after adjustment by initial hemoglobin values, and both groups were very similar. These data support the concept that the calcium-iron interaction under regular antenatal care does not appear to be clinically significant.

Iron and folate. Anemia, identified by several definitions, was the main outcome in the routine iron and folate supplementation reviews (90, 92) Hemorrhage-related outcomes were reported in calcium, magnesium and zinc supplementation reviews in addition to iron and folate.

Routine iron (12 trials [See Annex Table 1 (63) Trials 35–48]), folate (6 trials [See Annex Table 1 (63) Trials 37,38,40,41,54,55]) and iron and folate (6 trials [See Annex Table 1 (63) Trials 35,37,41,48,60,61]) supplementation results in a substantial reduction of women with hemoglobin levels <100 g/L in late pregnancy (Table 1). Other laboratory measures of iron and folate status and anemia are similarly affected by supplementation. The trial comparing routine with selective iron supplementation showed a reduction in the need for postpartum blood transfusion in the routine supplementation group [See Annex Table 1 (63) Trial 49]. The fact that the trial was not blinded, and the iron dose was not standardized for all participants, suggests that these findings should be interpreted with caution.

A randomized trial recently evaluated a new model of antenatal care; results support the protective effect on the incidence of postpartum anemia of actively promoting and providing routine iron and folate supplementation to all pregnant women (22). The compliance and absorption rate concerns with daily iron supplementation led investigators to compare daily versus less-frequent iron supplementation regimens. Three trials not yet included in the Cochrane systematic reviews compared daily versus weekly or twice weekly oral iron-folate supplementation (129– 131) and one trial compared daily iron-folate with four weekly iron dextran supplements (132). Mumtaz et al. (130) found that daily administration resulted in a greater rise in hemoglobin levels in Pakistani women whereas the other two trials conducted in Indonesia (129) and in Malawi (131) did not find any difference in the rise in hemoglobin levels. However, these latter two trials doubled the dose of iron and folate in the weekly supplemented groups. Beaton and McCabe (133) conducted a meta-analysis on the efficacy of weekly vs. daily iron supplementation during pregnancy. The meta-analysis included four trials with individual randomization, three trials with cluster randomization and one trial with no information on the randomization process. Overall, weekly dosing was less efficacious than daily supplementation (RR 1.29; 95%CI: 1.10–1.51). However, two of the eight trials showed a protective effect of weekly versus daily supplementation, and there was indication of heterogeneity in the trial results (²=11.06, 7 d.f.).

Magnesium. The Cochrane review of magnesium supplementation showed a statistically significant large reduction in antepartum hemorrhage (86). This result is based on two trials [See Annex Table 1 (63) Trials 67,68] but was strongly influenced by the result of the second trial that used quasi-random allocation. There is no likely biological explanation for this effect; and the authors of the review attribute this relationship to the poor quality of one trial and the high likelihood of bias.

Fish oil. The systematic review of fish oil supplementation is currently underway (134). Anemia- and hemorrhage-related outcomes are not the expected outcomes of this intervention although fish oil supplementation could, in theory, increase the risk of hemorrhage as a side effect.

Zinc. The Cochrane systematic review of clinical trials of zinc supplementation in pregnancy did not estimate differences in hemoglobin levels by supplement type. Two randomized trials published after the last update of the review compared hemoglobin levels in zinc-supplemented and non-zinc-supplemented women and did not observe differences by supplement type (135, See Annex Table 1 (63) Trial 74].

Vitamin A. Three randomized trials of vitamin A supplementation present anemia-related outcomes. The Cochrane systematic review of vitamin A supplementation in pregnancy has not been published and only a protocol is available (136). Suharno et al. (6) suggested that a combination of iron and vitamin A resulted in a greater rise in hemoglobin in Indonesia (6). Trials by Semba et al. [See Annex Table 1 (63) Trial 90] and Van den Broek et al. (N.R. van den Broek, S.A. White, J.D. Cook, E.A. Letsky, M. Molyneux and J.P. Neilson, unpublished, 2002), both conducted in Malawi in women receiving iron and folate supplements, did not find any improvement in the number of women with anemia.

    Nutritional interventions for the prevention of maternal infection. Nutritional advice, protein-energy, calcium, iron and folate and magnesium supplementation reviews did not report maternal infection as an outcome. Maternal infection was not considered as an outcome for fish oil interventions.

Zinc. In the zinc supplementation Cochrane review (91), one trial conducted in England reported on maternal infection [See Annex Table 1 (63) Trial 80]. Although this trial was of high methodological quality, it was conducted in an adequately nourished population and no differences in maternal infection rate were found (Table 1). None of the studies published after the last update of the systematic review reported maternal infection as an outcome.

Vitamin A. The Nepal placebo-controlled vitamin A and ß-carotene supplementation trial [See Annex Table 1 (63) Trial 92] found a significant reduction in all-cause maternal mortality (40% reduction with vitamin A and 50% reduction with ß-carotene). Maternal deaths due to infection were similar in the vitamin A and placebo groups but fewer in the ß-carotene group (Table 3). In the systematic review of vitamin A supplementation in pregnancy none of the other included trials reported data on maternal infection and related outcomes (136).

Numerous studies of healthy women from both affluent countries (137– 142) and less affluent countries (143– 145) have shown shorter maternal height and/or larger newborn weight to be associated with increased delivery complications, whereas an increase in fetal growth is associated with lower risk of perinatal distress. Conversely, smaller women tend to have smaller babies, many of whom will show signs of IUGR, raising the possibility that this biological relationship may protect shorter women from suffering an excess of delivery complications. A detailed discussion of this relationship has been published (146).

There are therefore two questions related to the evaluation of nutritional interventions on the rate of obstructed labor: 1) can nutritional interventions during pregnancy aimed at improving fetal growth increase the risk of obstructed labor (147) echoing the concerns that some pregnant women have expressed, or acted upon, "eating down" during pregnancy (148, 149) or, conversely, improving maternal nutritional status could reduce pregnancy complications, making mothers more able to go though labor, thus reducing the risk of cesarean section and/or obstructed labor? 2) can nutritional interventions aimed at optimizing growth during fetal life or childhood in girls, be associated with a reduction in the risk, later in adulthood, of obstructed labor?

An additional element in the understanding of these relationships is the definition of obstructed labor. In most of the trials examined, such outcome is not included and therefore the incidence of intrapartum cesarean delivery (or cesarean delivery overall) is used as a proxy indicator for obstructed labor (146).

Cesarean section, as an independent outcome, or combined with other indicators of obstructed labor was reported in the Cochrane systematic reviews of balanced protein-energy supplementation, calcium, iron, folate, magnesium, fish oil, and zinc and is used here as a proxy for obstructed labor.

Balanced protein-energy supplementation. None of the trials included in the Cochrane review (81) presented information on cesarean section. One trial, however, presented information on duration of labor which was similar in the two groups, 6.20 (±5.60) hours in the supplemented group versus 6.30 (±4.70) hours in the control group [See Annex Table 1 (63) Trial 12] (Table 1).

Salt restriction. The Cochrane review of reduced salt intake compared with normal dietary salt, or high intake, in pregnancy includes only one trial in Holland [See Annex Table 1 (63) Trial 22] that reported data on cesarean section rates (84). The results of this trial show no statistically significant evidence of an effect of salt restriction in pregnancy on cesarean section rates (Table 1).

Calcium. The Cochrane review on calcium (86) found that the rate of cesarean section was similar in the supplemented and placebo groups, even after stratifying by calcium intake level (inadequate and adequate), and by risk of developing hypertension in pregnancy, with the exception of women at high risk of hypertension who had a trend toward lower cesarean section in the calcium supplemented group (Table 1 [See Annex Table 1 (63) Trials 24–26, 31–33]).

Iron and folic acid. The systematic review for iron in pregnancy (92) reported a statistically significant increase in the cesarean section rate associated with routine iron supplementation compared with selective iron in the only trial included in this review [See Annex Table 1 (63) Trial 49]. In addition, the systematic review of folic acid and iron supplementation in pregnancy (89) reported a nonstatistically significant reduction in cesarean section rate with iron and folic acid versus no iron and folic acid although these results were based on two very small trials [See Annex Table 1 (63) Trials 35,52] (Table 1).

There is one systematic review evaluating treatments for iron deficiency anemia in pregnancy (93). One small trial [See Annex Table 1 (63) Trial 62], included in this review, compared the rates of cesarean section between intravenous iron versus regular oral iron and found a non statistically significant reduction in the rate of cesarean section associated with a regular oral iron treatment regimen. One review of folate supplementation in pregnancy (90) included two trials [See Annex Table 1 (63) Trials 51,52] showing a non significant reduction in cesarean section rates (Table 1).

Magnesium. In the systematic review of magnesium supplementation in pregnancy (87) only one trial [See Annex Table 1 (63) Trial 68] reported results on the length of labor, which were similar between groups (mean length 6 h in both magnesium supplementation and placebo groups).

Zinc. The Cochrane review of zinc supplementation in pregnancy (91), which included three trials [See Annex Table 1 (63) Trials 78,80,85] with rates of cesarean section, detected a significant preventive effect of zinc supplementation on cesarean section rates (Table 1). Studies published after the update of the review do not report obstructed labor or other cesarean section related outcomes [See Annex Table 1 (63) Trials 74,82]. We do not have data on obstructed labor from unpublished studies. One study conducted in West Java, published as doctoral thesis, reported an increase in delivery complications among the zinc supplemented without providing details of the type of complications [See Annex Table 1 (63) Trial 76].

Fish oil. Presently no data are available on obstructed labor related outcomes from the review on fish oil supplementation in pregnancy. However, a protocol for a new systematic review has been submitted to and accepted by the Cochrane Library (134).

Implications of nutritional interventions to increase birth weight. Data to evaluate the effect of interventions aimed at improving fetal growth were obtained from systematic reviews of randomized controlled trials (82, 149, 150). The results from 11 randomized controlled trials (151– 155, See Annex Table 1 (63) Trials 5,7,8,12,14,16) that included 3526 women showed that balanced protein/energy supplementation (<25% of total energy as protein) was associated with a nonstatistically significant mean increase of 25 g in the mean birth weight (95% CI of the mean difference -3.6 g to 54.5 g). Five of the 11 trials, which included 1507 women, showed a nonstatistically significant increase in mean newborn head circumference of 0.07 cm (95% CI of the difference -0.06 cm to 0.20 cm) associated with supplementation (82). This very small change in head circumference is similar to that reported in a trial from The Gambia (156) that was not included for this outcome in the systematic review (82). Therefore, there is no evidence that balanced energy/protein supplementation during pregnancy results in a biologically meaningful difference on the head circumference of the fetus and newborn.

Using regression equations, we recently calculated the effect of an increase in birth weight of 100 g on the risk of cesarean delivery and perinatal distress across the cutoff for low birth weight (2500 g) (146). To make the estimations, specific values of birth weight have to be chosen because the change in risk is not uniform across the birth weight distributions. If an intervention changes birth weight from 2450 g (low birth weight) to 2550 g for 1000 newborns, it is estimated that an additional eight infants will be delivered by intrapartum cesarean section. Correspondingly, 34 fewer cases of perinatal distress would occur in a group of 1000 newborns with this increase in birth weight. Using a much larger theoretical increase of 600 g in birth weight, an unlikely outcome from any antenatal care intervention, from a low birth weight of 2460–3060 g, then for every 1000 cases experiencing this change there would be 51 additional intrapartum cesarean deliveries and 123 fewer cases of perinatal distress.

Finally there is evidence from the supplementation studies in Guatemala that, when offered ad libitum, less supplement is consumed by women who have better nutritional status and lower reproductive stress (157). This supports the contention that maternal nutritional supplementation during pregnancy among chronically nutritionally deprived populations is unlikely to result in macrosomia.

Increase in maternal height during early life. In a longitudinal follow-up study in Guatemala, girls who has been nutritionally supplemented in early childhood attained an average of 1.5 cm increase in height above the mean of their mothers by the age of 18–25 years (158). Again, using a prediction equation, this actual increase in height attributed to nutritional supplementation was estimated to result in a 2.2% decrease in their risk of intrapartum cesarean if they gave birth to a child weighing 3060 g. Another useful estimation can be generated by calculating the probability of an intrapartum cesarean delivery if an increase in height was achieved among the Guatemalan women compared with the mean increase in height of American-born women of Mexican descent, i.e., an increase in height from 153 cm to 158 cm (159). The decrease in risk of intrapartum cesarean delivery would be from approximately 24% to 17% (assuming the same maternal age, newborn size and age and no perinatal distress).

    Nutritional interventions for the prevention of maternal mortality. None of the trials or systematic reviews of nutritional interventions during pregnancy addressed maternal mortality in a satisfactory way. Some reviews identified mortality as a relevant outcome but, except for the vitamin A supplementation review, all had very few outcome events in this respect.

The systematic review on the effects of vitamin A on pregnancy outcome includes data from the large cluster-randomized trial conducted in Nepal by West et al. [See Annex Table 1 (63) Trial 92] that evaluated the effect of weekly low-dose vitamin A or ß-carotene. This trial reported a reduction in all-cause maternal mortality up to 12 wk postpartum in the supplemented groups compared with the placebo group (Table 3). Stratified analyses were conducted by timing and cause of death, but these analyses did not have the power to reach statistically significant levels. The effect was consistent across strata by timing of death. The effect by cause of death is not consistent and appears to be mostly concentrated in preventing deaths related to obstetric, infection, injury and miscellaneous causes due to ß-carotene supplementation compared with placebo. A vitamin A effect was only observed in the miscellaneous stratum. Thus, these data should not be presented as an overall effect. For miscellaneous causes, including chronic illnesses and uncertain causes, there were fewer statistically significant maternal deaths in the vitamin A group and the ß-carotene group compared with the placebo group. Although the study was large and community based, the number of events was not sufficient to give robust results in the stratified analysis. The results, however, suggest ß-carotene supplementation may have an impact on maternal mortality. In populations with vitamin A deficiency, programs to increase vitamin A or ß-carotene intake must be initiated.

    Nutritional interventions for the prevention of preterm delivery. Nutritional advice. Of the four trials included in the systematic review (81), only one reported on preterm delivery rates [See Annex Table 1 (63) Trial 2]. Although the trial shows a protective effect of the nutritional advice on preterm deliveries, the conclusions are limited because of many post randomization exclusions and because this cluster-randomized trial was analyzed as individual randomization ignoring the cluster effect (Table 4).

Three trials had isocaloric protein-energy supplementation (83); only one, of poor methodological quality [See Annex Table 1 (63) Trial 11], reported rate of preterm delivery as an outcome showing no differences between the supplemented and control groups (Table 4). Two trials included women receiving high-protein supplementation (160). These trials assessed the effects of >25% of the total energy supplementation from protein. Only one trial reported the effect on preterm birth [See Annex Table 1 (63) Trial 15]. This methodologically sound trial showed that high protein supplementation was not associated with a clinically or statistically difference in the rate of preterm deliveries (Table 4).

Protein and energy restriction. Three trials were included in a systematic review (84). Only one small trial reported results on preterm deliveries [See Annex Table 1 (63) Trial 21]. This trial showed a reduction in the rate of preterm deliveries in the intervention group compared with the control group, but the CI were very wide and compatible with both a protective and a deleterious effect (Table 4). These results should be interpreted cautiously because the treatment allocation method was not reported and the reporting of gestational age was unclear.

Salt restriction. Of the two randomized controlled trials included in the review (85), only one reported on preterm deliveries [See Annex Table 1 (63) Trial 23a]. The small sample size of this trial, however, does not allow reliable answers to be extracted (Table 4).

Calcium. A previous meta-analysis of five calcium supplementation trials with preterm birth outcomes showed a reduced rate of preterm delivery (typical RR 0.69; 95% CI: 0.48–1.01) (161). Since that publication, four new randomized double-blind controlled trials of calcium supplementation have been published. Of these, three had clinical outcomes [See Annex Table 1 (63) Trials 26,29,31] and one focused only on the hemodynamic changes during the third trimester of pregnancy (162).

Two of the new clinical trials were conducted in calcium-deficient populations in Ecuador [See Annex Table 1 (63) Trial 29] and India [See Annex Table 1 (63) Trial 31]; the large multicenter National Institutes of Health trial was conducted in healthy nulliparous women with normal baseline calcium intake (mean 1130 g/d) in the United States [See Annex Table 1 (63) Trial 26]. The U.S. trial did not demonstrate any protective effect of calcium supplementation on preterm (<37 wk) or very preterm (<34 wk) births or an increase in the duration of gestation. The trial from India enrolled women with an average calcium intake of 350 mg/d. The preterm delivery rates were 2.06% in the calcium supplementation group and 6.45% in the placebo group (RR 0.32; 95% CI: 0.07–1.54). The Ecuador trial was conducted in teenagers with a baseline calcium intake of approximately 600 mg/d. This trial demonstrated a statistically significant increase in the mean gestational age from 38.7 wk (standard error 0.3) in the placebo group to 39.6 wk (standard error 0.4) in the calcium-supplemented group.

We conducted a new meta-analysis following the strategy used in the recently published Cochrane review (86). Only trials with random allocation, double blind and placebo-controlled methodology with adequate allocation concealment were included. Analyses were stratified according to two prespecified criteria: risk of hypertensive disorders of pregnancy and baseline calcium intake. There were nine trials for which preterm birth rates could be estimated [See Annex Table 1 (63) Trials 24,26,26,28,29,31,32,33,34]. Overall, no protective effect of calcium supplementation on preterm delivery could be demonstrated (Table 4). The meta-analysis stratified by women who were at risk of high blood pressure in pregnancy included 270 women in the calcium group and 298 women in the placebo group in four trials with a typical RR 0.43; 95% CI: 0.23–0.78. Five trials studying populations with low calcium intake reported preterm delivery rates (799 women in the calcium group and 838 in placebo group). There was a statistically non significant reduction in preterm delivery rates with calcium supplementation (typical RR 0.77; 95% CI: 0.52–1.14). No protective effect of calcium supplementation on preterm delivery was observed among women at low-risk of pregnancy hypertension or adequate dietary calcium intake. These results are in agreement with the most recent update of the Cochrane Library systematic review (86).

Iron and folate. The Cochrane systematic review of iron supplementation contains 20 trials (92). Only one trial reported effects on preterm births [See Annex Table 1 (63) Trial 49]. This large trial was conducted in a well-nourished Finish population; routine iron supplementation compared with selective supplementation indicated no difference between groups, although there appeared to be a trend toward an increase risk with iron (Table 4).

Twenty-one studies were included in the review of routine folate supplementation alone (90), but only four reported the effect on preterm delivery [See Annex Table 1 (63) Trials 51,52,53,55]. The group with folate supplementation alone during pregnancy did not show any difference from the control group (Table 4).

Eight trials were included in the review of iron and folate supplementation in pregnancy (89). Only one small trial reported the effect of iron and folate supplementation on preterm deliveries [See Annex Table 1 (63) Trial 61]. Although the results showed a detrimental effect of iron and folate supplementation on preterm delivery, the trial was very small, with a wide CI compatible with both a protective and a deleterious effect.

Magnesium. The Cochrane review on magnesium supplementation (87) was recently updated and comprises seven trials, five of which reported preterm birth as an outcome [See Annex Table 1 (63) Trials 63,65–68]. Oral magnesium supplementation started before 25 wk gestation was associated with a statistically significant reduction in preterm birth rates (Table 4). However, these results should be interpreted cautiously because most of these trials had methodological weaknesses and the trial with the highest methodological quality [See Annex Table 1 (63) Trial 67] showed no effect on reducing preterm birth. The heterogeneity among trial results seems to be related to the methodological quality of the included trials, with less protective effect observed in large trials or those with higher quality.

Fish oil. The fish oil review was published 8 y ago (88) and included three trials of which two reported preterm delivery [See Annex Table 1 (63) Trials 72,70b]. The first trial dominated the review because of its large sample size (N = 5017); the experimental group in this trial received vitamins and minerals as well as fish oil. The results indicate a statistically significant 20% reduction in preterm birth rates with fish oil supplementation with narrow CIs. However, the trial has methodological limitations, such as the alternate treatment allocation method and no outcome measurement blinding. The second trial included in the review is smaller (N = 513) that the first but methodologically sound.

Three other published articles were identified. The Onwude et al. [See Annex Table 1 (63) Trial 71] randomized double-blind controlled trial of fish oil versus placebo showed no differential effect on preterm deliveries (Table 2). However, the CI is wide reflecting that the sample size was inadequate for drawing any reliable conclusions. The Bulstra-Ramkers et al. [See Annex Table 1 (63) Trial 69] randomized double-blind controlled trial reported that fish oil had no different effect from placebo on preterm delivery (Table 2). Again the number of women included was small. Recently six other trials were published as part of an international multicenter program [See Annex Table 1 (63) Trial 70a]. Among the trials included in this non-Cochrane review, one used fish oil prophylactically—among women who had previously experienced preterm delivery—and showed a statistically significant reduction in the recurrence of preterm delivery compared with olive oil (Table 2). The results of the trials are not conclusive but fish oil appears to be a promising intervention for preventing preterm delivery. The evidence extracted from the reviewed trials supports the need for a large, well-designed randomized controlled trial for a definitive answer about whether fish oil supplementation is effective in reducing preterm delivery.

Zinc. The systematic review of clinical trials of zinc supplementation included seven trials (91), 5 with preterm delivery data [Annex Table 1 (63) Trials 77–80,85]. Overall, zinc supplementation was protective (Table 4). In one trial [See Annex Table 1 (63) Trial 77], only women with low plasma zinc levels were eligible. This trial showed statistically significant birth weight increases in infants of women with body mass index less than 26, although there were no statistically significant differences in preterm deliveries defined as <37 or <33 wk.

None of the studies published after the last update of the review observed an effect of zinc supplementation on gestational age or rate of preterm delivery (Table 5).

    Other outcomes. Nausea and vomiting of pregnancy. Evidence from one trial in 997 women showed that women taking periconceptional multivitamins were less likely to have severe nausea and vomiting in pregnancy (RR 0.46; 95% CI 0.26–0.79) (163, 164). A few nutritional interventions (vitamin B-6, thiamin, vitamin B-12) have been suggested to be effective and safe for the prevention of these conditions (165).

Maternal night vision. A study conducted in Nepal [See Annex Table 1 (63) Trial 75] showed that zinc supplementation augmented the effect of vitamin A in restoring night vision, a functional symptom of vitamin A deficiency in pregnant women.

Fetal neurobehavioral development. Two studies conducted in Peru found evidence of improved fetal neurobehavioral development, assessed by examining the changes over time in patterns of fetal heart rate and fetal movements, in fetuses of zinc-supplemented women [(51, 166), See Annex Table 1 (63) Trial 81]. These studies were conducted by the same investigators in similar populations. However, the consistency of the results, together with the positive effects detected by ultrasonography on fetal femur length growth, suggests that nutritional interventions during pregnancy may have effects that are subtle but still with important implications for postnatal life.

Discussion

Based on the available data from systematic reviews of randomized controlled trials up to July 2002, we can conclude that limited evidence supports large-scale interventions multivitamin, mineral or protein-energy to prevent or treat hypertensive disorders of pregnancy, obstructed labor or intrapartum cesarean section (during the index pregnancy), maternal infection, hemorrhage and preterm delivery. A large trial of good methodological quality, in women with low vitamin A intakes, observed that ß-carotene and vitamin A supplementation reduced maternal mortality.

Calcium supplementation appears promising for women with low calcium intakes who are at high risk for preeclampsia and preterm delivery. The results of an ongoing WHO calcium supplementation trial should confirm or reject this hypothesis. Fish oil and magnesium supplementation are also promising for preventing preterm delivery. There is a suggestion that oxidative stress and the activation of the inflammatory response, associated with reduced placental perfusion, have a role in the etiology of preeclampsia. Three trials of antioxidant therapy, presently being prepared, should answer several related questions.

In some instances, relative risks and confidence values of similar magnitude are associated with different levels of strength in our recommendations. This is related to the critical appraisal of the trials, which includes post randomization exclusions; heterogeneity of trials' results, with larger effect detected by small trials and smaller effect found in large trials; an imperfect system of treatment allocation concealment; and the addition of multivitamins to the control groups.

We have included all systematic reviews and randomized trials up to July 2002 as part of a WHO major effort to update and disseminate evidence-based information on maternal health care. One example is the free distribution of the WHO Reproductive Health Library, a CD ROM presenting Cochrane systematic reviews and useful information on reproductive health problems to health workers in developing countries (167).

Pregnant women in developing countries have several nutrient deficiencies that have to be corrected to improve their health status and perinatal outcomes. Iron and folic acid supplementation are effective for preventing or treating severe anemia even postpartum. In interpreting results it is important to consider that some of the nutrition interventions have been evaluated in populations that are not nutritionally deficient (pharmacological effect). We highlighted instances when we identified differential effects by baseline characteristics. The preventive effect of calcium supplementation on preeclampsia in women with low calcium intake is an example of this situation (Table 1). Although the proportion of trials conducted in populations at risk for nutritional deficiencies varied across systematic reviews and outcomes, most reviews included trials conducted in populations at risk of nutritional deficiency (Tables 2 and 5). An exception is the effect of iron supplementation, which has been studied mostly in populations from developed countries. Overall, we feel that the general conclusion of this overview, with its heterogeneity, can be reasonably applied to populations at different levels of nutritional status.

In this review, we have focused on randomized clinical trials because they are most likely to provide unbiased results, which is particularly important in research on nutritional interventions, where many variables can bias results. As we previously discussed (149), a worrisome gap exists between the magnitude of the problems (including maternal nutritional status and negative pregnancy outcomes), logistics needed for the implementation and cost of nutritional interventions and the overall quality and sample size of the randomized controlled trials that evaluated such nutritional interventions. The scientific community, donors and multilateral organizations should make efforts to correct these major barriers to effective health programs.

For future research efforts, the best strategy would be to continue to test promising interventions in populations with and without nutritional deficiency in large multicenter trials that require short time periods (168). However, it should be put in scientific context that for most of pregnancy-specific conditions such as preeclampsia, postpartum hemorrhage (which, when defined as losses >1000 mL, still has an incidence of 3–5% of deliveries), preterm delivery, premature rupture of the membranes and initiation of labor, we have limited knowledge of their etiology. Therapeutic options for these conditions remain limited to pregnancy termination or palliative treatments to be started when the condition is already manifest.

We believe that only a major coordinated research effort based on a systematic review of the literature that is focused and well funded can address these problems. WHO has started such a program in relation to the etiology and pathophysiology of preeclampsia (169).The promising areas identified by the systematic reviews will guide the implementation of collaborative research efforts from basic science studies to randomized clinical trials to explain and test the identified hypotheses.

When planning and implementing program interventions, one should be aware that a few months of nutrient supplementation during pregnancy may be not be sufficient to offset decades of low nutrient intake and its adaptive clinical mechanisms. After all, as Sir Douglas Bird put it: "There is, of course, a limit to what can be achieved in one generation in improving the level of health and physique" (170).

	FOOTNOTES

 
¹ Manuscript prepared for the USAID-Wellcome Trust workshop on "Nutrition as a preventive strategy against adverse pregnancy outcomes," held at Merton College, Oxford, July 18–19, 2002. The proceedings of this workshop are published as a supplement to The Journal of Nutrition. The workshop was sponsored by the United States Agency for International Development and The Wellcome Trust, UK. USAID's support came through the cooperative agreement managed by the International Life Sciences Institute Research Foundation. Supplement guest editors were Zulfiqar A. Bhutta, Aga Khan University, Pakistan, Alan Jackson (Chair), University of Southampton, England, and Pisake Lumbiganon, Khon Kaen University, Thailand.

² The views expressed in this document are solely the responsibility of the authors and do not necessarily represent the views of the World Health Organization or its member states.

⁴ Abbreviations: CI, confidence interval; IUGR, intrauterine growth restriction; RR, relative risk; WHO, World Health Organization.

	LITERATURE CITED

TOP ABSTRACT LITERATURE CITED

 

1. Villar, J., Belizan, J. M. & Fischer, P. J. (1983) Epidemiologic observations on the relationship between calcium intake and eclampsia. Int. J. Gynaecol. Obstet. 21: 271–278.[Medline]

2. DeCherney, A. H. & Koos, B. (1997) Obstetrics and gynecology. JAMA 277: 1878–1879.[Free Full Text]

3. de Silva, L. D. & Atukorala, T. M. (1996) Micronutrient status of plantation workers in Sri Lanka during pregnancy and postpartum. J. Obstet. Gynaecol. Res. 22: 239–246.[Medline]

4. Katz, J., Khatry, S. K., West, K. P., Humphrey, J. H., Leclerq, S. C., Kimbrough, E., Pohkrel, P. R. & Sommer, A. (1995) Night blindness is prevalent during pregnancy and lactation in rural Nepal. J. Nutr. 125: 2122–2127.

5. Llewellyn-Jones, D. (1990) Fundamentals of Obstetrics and Gynaecology, Vol. 1 Mosby-Year Book, London.

6. Suharno, D., West, C. E., Muhilal, Karyadi, D. & Hautvast, J. G. (1993) Supplementation with vitamin A and iron for nutritional anaemia in pregnant women in West Java. Indonesia. Lancet 342: 1325–1328.

7. West, K. P., Jr., Katz, J., Khatry, S. K., Leclerq, S. C., Pradhan, E. K., Shrestha, S. R., Connor, P. B., Dali, S. M., Christian, P., Pokhrel, R. P. & Sommer, A. (1999) Double blind, cluster randomised trial of low dose supplementation with vitamin A or beta carotene on mortality related to pregnancy in Nepal. The NNIPS–2 Study Group. BMJ 318: 570–575.[Abstract/Free Full Text]

8. Vallee, B. L. & Falchuck, K. H. (1993) The biochemical basis of zinc physiology. Physiol. Rev. 73: 79–118.[Free Full Text]

9. Prasad, A. S. (1996) Zinc deficiency in women, infants and children. J. Am. Coll. Nutr. 15: 113–120.[Abstract]

10. Jameson, S. (1993) Zinc status in pregnancy: the effect of zinc therapy on perinatal mortality, prematurity, and placental ablation. Ann. N. Y. Acad. Sci. 678: 178–192.[Medline]

11. Villar, J., Ezcurra, E., Gurtner de lau Fuente, V. & Campodonico, L. (1994) Pre-term delivery syndrome: the unmet need In New Perspectives for the effective treatment of pre-term labor: an international consensus. Research and Clinical Forums 16: 9–38.

12. McGaw, T. (2002) Periodontal disease and preterm delivery of low-birth-weight infants. J. Can. Dent. Assoc. 68: 165–169.

13. Romero, R., Oyarzun, E., Mazor, M., Sirtori, M., Hobbins, J. C. & Bracken, M. (1989) Meta-analysis of the relationship between asymptomatic bacteriuria and preterm delivery/low birth weight. Obstet. Gynecol. 73: 576–582.[Medline]

14. Goldenberg, R. L. & Andrews, W. W. (1996) Intrauterine infection and why preterm prevention programs have failed. Am. J. Public Health 86: 781–783.[Free Full Text]

15. Taha, T. T., Gray, R. H. & Mohamedani, A. A. (1993) Malaria and low birth weight in central Sudan. Am. J. Epidemiol. 138: 318–325.[Abstract/Free Full Text]

16. Kramer, M. S. (2003) Maternal nutrition and adverse pregnacy outcome: an overview. J. Nutr. 133: 1592S–1596S.

17. Fall, C. H. D., Yajnik, C. S., Rao, S., Davies, A. A., Brown, N. & Farrant, H. J. W. (2003) A review of the evidence that maternal micronutrient status influences fetal growth and survival. J. Nutr. 133: 1747S–1756S.[Abstract/Free Full Text]

18. Villar, J. & Carroli, G. (1996) Methodological issues of randomized controlled trials for the evaluation of reproductive health interventions. Prev. Med. 25: 365–375.[Medline]

19. Klebanoff, M. A., Carey, J. C., Hauth, J. C., Hillier, S. L., Nugent, R. P., Thom, E. A., Ernest, J. M., Heine, R. P., Wapner, R. J., Trout, W., Moawad, A., Leveno, K. J., Miodovnik, M., Sibai, B. M., Van Dorsten, J. P., Dombrowski, M. P., O'Sullivan, M. J., Varner, M., Langer, O., McNellis, D. & Roberts, J. M. (2001) Failure of metronidazole to prevent preterm delivery among pregnant women with asymptomatic Trichomonas vaginalis infection. N. Engl. J. Med. 345: 487–493.[Abstract/Free Full Text]

20. Carey, J. C., Klebanoff, M. A., Hauth, J. C., Hillier, S. L., Thom, E. A., Ernest, J. M., Heine, R. P., Nugent, R. P., Fischer, M. L., Leveno, K. J., Wapner, R. & Varner, M. (2000) Metronidazole to prevent preterm delivery in pregnant women with asymptomatic bacterial vaginosis. N. Engl. J. Med. 342: 534–540.[Abstract/Free Full Text]

21. Klebanoff, M. A., Regan, J. A., Rao, A. V., Nugent, R. P., Blackwelder, W. C., Eschenbach, D. A., Pastorek, J. G., Williams, S., Gibbs, R. S. & Carey, J. C. (1995) Outcome of the vaginal infections and prematurity study: results of a clinical trial of erythromycin among pregnant women colonized with group B streptococci. Am. J. Obstet. Gynecol. 172: 1540–1545.[Medline]

22. Villar, J., Ba'aqeel, H., Piaggio, G., Lumbiganon, P., Miguel, B. J., Farnot, U., Al Mazrou, Y., Carroli, G., Pinol, A., Donner, A., Langer, A., Nigenda, G., Mugford, M., Fox-Rushby, J., Hutton, G., Bergsjo, P., Bakketeig, L., Berendes, H. & Garcia, J. (2001) WHO antenatal care randomised trial for the evaluation of a new model of routine antenatal care. Lancet 357: 1551–1564.[Medline]

23. Carroli, G., Villar, J., Piaggio, G., Khan-Neelofur, D., Gulmezoglu, M., Mugford, M., Lumbiganon, P., Farnot, U. & Bersgjo, P. (2001) WHO systematic review of randomised controlled trials of routine antenatal care. Lancet 357: 1565–1570.[Medline]

24. Villar, J. & Belizan, J. M. (1982) The timing factor in the pathophysiology of the intrauterine growth retardation syndrome. Obstet. Gynecol. Surv. 37: 499–506.[Medline]

25. Goldstein, I., Lockwood, C. & Hobbins, J. C. (1987) Ultrasound assessment of fetal intestinal development in the evaluation of gestational age. Obstet. Gynecol. 70: 682–686.[Medline]

26. Murao, F., Takamori, H., Hata, K., Hata, T. & Kitao, M. (1987) Fetal liver measurements by ultrasonography. Int. J. Gynaecol. Obstet. 25: 381–385.[Medline]

27. Duncan, K. R., Gowland, P. A., Moore, R. J., Baker, P. N. & Johnson, I. R. (1999) Assessment of fetal lung growth in utero with echo-planar MR imaging. Radiology 210: 197–200.[Abstract/Free Full Text]

28. Garden, A. S. & Roberts, N. (1996) Fetal and fetal organ volume estimations with magnetic resonance imaging. Am. J. Obstet. Gynecol. 175: 442–448.[Medline]

29. WHO Expert Committee on Physical Status. (1995) The newborn infant. In: Physical Status: The use and interpretation of anthropometry, pp. 121–160. WHO, Geneva.

30. Campbell, S. & Thoms, A. (1977) Ultrasound measurement of the fetal head to abdomen circumference ratio in the assessment of growth retardation. Br. J. Obstet. Gynaecol. 84: 165–174.[Medline]

31. Harding, J. E. (2001) The nutritional basis of the fetal origins of adult disease. Int. J. Epidemiol. 30: 15–23.[Free Full Text]

32. Cuttini, M., Cortinovis, I., Bossi, A. & de Vonderweid, U. (1991) Proportionality of small for gestational age babies as a predictor of neonatal mortality and morbidity. Paediatr. Perinat. Epidemiol. 5: 56–63.[Medline]

33. Villar, J., de Onis, M., Kestler, E., Bolanos, F., Cerezo, R. & Bernedes, H. (1990) The differential neonatal morbidity of the intrauterine growth retardation syndrome. Am. J. Obstet. Gynecol. 163: 151–157.[Medline]

34. Caulfield, L. E., Haas, J. D., Belizan, J. M., Rasmussen, K. M. & Edmonston, B. (1991) Differences in early postnatal morbidity risk by pattern of fetal growth in Argentina. Paediatr. Perinat. Epidemiol. 5: 263–275.[Medline]

35. Villar, J., Cogswell, M., Kestler, E., Castillo, P., Menendez, R. & Repke, J. T. (1992) Effect of fat and fat-free mass deposition during pregnancy on birth weight. Am. J. Obstet. Gynecol. 167: 1344–1352.[Medline]

36. Lederman, S. A., Paxton, A., Heymsfield, S. B., Wang, J., Thornton, J. & Pierson, R. N., Jr. (1999) Maternal body fat and water during pregnancy: do they raise infant birth weight? Am. J. Obstet. Gynecol. 180: 235–240.[Medline]

37. Mardones-Santander, F., Salazar, G., Rosso, P. & Villarroel, L. (1998) Maternal body composition near term and birth weight. Obstet. Gynecol. 91: 873–877.[Medline]

38. Deleted in proof.

39. Villar, J. & Rivera, J. (1988) Nutritional supplementation during two consecutive pregnancies and the interim lactation period: effect on birth weight. Pediatrics 81: 51–57.[Abstract/Free Full Text]

40. Levine, R. J., Hauth, J. C., Curet, L. B., Sibai, B. M., Catalano, P. M., Morris, C. D., DerSimonian, R., Esterlitz, J. R., Raymond, E. G., Bild, D. E., Clemens, J. D. & Cutler, J. A. (1997) Trial of calcium to prevent preeclampsia. N. Engl. J. Med. 337: 69–76.[Abstract/Free Full Text]

41. Belizan, J. M. & Villar, J. (1980) The relationship between calcium intake and edema-, proteinuria-, and hypertension-getosis: an hypothesis. Am. J. Clin. Nutr. 33: 2202–2210.[Free Full Text]

42. Belizan, J. M., Villar, J., Gonzalez, L., Campodonico, L. & Bergel, E. (1991) Calcium supplementation to prevent hypertensive disorders of pregnancy. N. Engl. J. Med. 325: 1399–1405.[Abstract]

43. Subcommittee on the Tenth Edition of the RDAs. (1989) Recommended Dietary Allowances, pp. 180. Food and Nutrition Board. Commission on Life Sciences. National Research Council National Academy Press, Washington, DC.

44. Rush, D., Stein, Z. & Susser, M. (1980) A randomized controlled trial of prenatal nutritional supplementation in New York City. Pediatrics 65: 683–697.[Abstract/Free Full Text]

45. Chappell, L. C., Seed, P. T., Briley, A. L., Kelly, F. J., Lee, R., Hunt, B. J., Parmar, K., Bewley, S. J., Shennan, A. H., Steer, P. J. & Poston, L. (1999) Effect of antioxidants on the occurrence of pre-eclampsia in women at increased risk: a randomised trial. Lancet 354: 810–816.[Medline]

46. Sebire, N. J., Jolly, M., Harris, J., Regan, L. & Robinson, S. (2001) Is maternal underweight really a risk factor for adverse pregnancy outcome? A population-based study in London. BJOG. 108: 61–66.[Medline]

47. The Magpie Trial Collaborative Group. (2002) Do women with preeclampsia, and their babies, benefits from magnesium sulphate? The Magpie Trial: a randomised placebo-controlled trial. Lancet 359: 1877–1890.[Medline]

48. The Collaborative Eclampsia Trial. (1995) Which anticonvulsant for women with eclampsia? Evidence from the Collaborative Eclampsia Trial. Lancet 345: 1455–1463.[Medline]

49. von Dadelszen, P., Ornstein, M. P., Bull, S. B., Logan, A. G., Koren, G. & Magee, L. A. (2000) Fall in mean arterial pressure and fetal growth restriction in pregnancy hypertension: a meta-analysis. Lancet 355: 87–92.[Medline]

50. Abalos, E., Duley, L., Steyn, D. W. & Henderson-Smart, D. J. (2002) Antihypertensive drug therapy for mild to moderate hypertension during pregnancy. Database Syst. Rev. CD002252.

51. Merialdi, M. (2001) The effect of maternal zinc supplementation during pregnancy on fetal growth and neurobehavioral development. PhD thesis, Johns Hopkins University, Baltimore, MD.

52. Godfrey, K. M. & Barker, D. J. (2000) Fetal nutrition and adult disease. Am. J. Clin. Nutr. 71: 1344S–1352S.[Abstract/Free Full Text]

53. Joseph, K. S. & Kramer, M. S. (1996) Review of the evidence on fetal and early childhood antecedents of adult chronic disease. Epidemiol. Rev. 18: 158–174.[Free Full Text]

54. Belizan, J. M., Villar, J., Bergel, E., del Pino, A., Di Fulvio, S., Galliano, S. V. & Kattan, C. (1997) Long-term effect of calcium supplementation during pregnancy on the blood pressure of offspring: follow up of a randomised controlled trial. BMJ 315: 281–285.[Abstract/Free Full Text]

55. Singhal, A., Cole, T. J. & Lucas, A. (2001) Early nutrition in preterm infants and later blood pressure: two cohorts after randomised trials. Lancet 357: 413–419.[Medline]

56. Villar, J. & Belizan, J. M. (1982) The relative contribution of prematurity and fetal growth retardation to low birth weight in developing and developed societies. Am. J. Obstet. Gynecol. 143: 793–798.[Medline]

57. Kramer, M. S. (1987) Determinants of low birth weight: methodological assessment and meta-analysis. Bull. World Health Organ. 65: 663–737.[Medline]

58. Goldenberg, R. L. & Rouse, D. J. (1998) Prevention of premature birth. N. Engl. J. Med. 339: 313–320.[Free Full Text]

59. Roberts, J. (2002) President's welcome. Hypertens. Pregnancy 21: vii–viii.

60. Breslow, N. E. & Day, N. E. (1980) Statistical methods in cancer research. Volume I—The analysis of case-control studies, pp 5–338. IARC Scientific Publications, Lyon.

61. Villar, J., Mackey, M. E., Carroli, G. & Donner, A. (2001) Meta-analyses in systematic reviews of randomized controlled trials in perinatal medicine: comparison of fixed and random effects models. Stat. Med. 20: 3635–3647.[Medline]

62. Villar, J., Piaggio, G., Carroli, G. & Donner, A. (1997) Factors affecting the comparability of meta-analyses and largest trials results in perinatology. J. Clin. Epidemiol. 50: 997–1002.[Medline]

63. Villar, J., Merialdi, M., Gülmezoglu, A. M., Abalos, E., Carroli, G., Kulier, R. & de Onis, M. (2003) Characteristics of randomized controlled trials included in systematic reviews of nutritional interventions reporting maternal morbidity, mortality, preterm delivery, intra-uterine growth restriction & small for gestational age and birth weight outcomes. J. Nutr. 133: 1632S–1639S[Free Full Text]

64. Clarke, M. & Oxman, A. D. Review Manager. &Oxman, A. D. Review Manager. (1999) Cochrane Reviewers' Handbook. (4.0): The Cochrane Collaboration. Oxford, England.

65. National Education Program Working Group on High Blood Pressure in Pregnancy. (2000) Report of the National High Blood Pressure Education Program Working Group on High Blood Pressure in Pregnancy. Am. J. Obstet. Gynecol. 183 (Suppl.), S1–S22.

66. Higgins, J. R. & de Swiet, M. (2001) Blood-pressure measurement and classification in pregnancy. Lancet 357: 131–135.[Medline]

67. Repke, J. T., Villar, J., Anderson, C., Pareja, G., Dubin, N. & Belizan, J. M. (1989) Biochemical changes associated with blood pressure reduction induced by calcium supplementation during pregnancy. Am. J. Obstet. Gynecol. 160: 684–690.[Medline]

68. Belizan, J. M., Villar, J. & Repke, J. (1988) The relationship between calcium intake and pregnancy-induced hypertension: up-to-date evidence. Am. J. Obstet. Gynecol. 158: 898–902.[Medline]

69. August, P., Marcaccio, B., Gertner, J. M., Druzin, M. L., Resnick, L. M. & Laragh, J. H. (1992) Abnormal 1,25–dihydroxyvitamin D metabolism in preeclampsia. Am. J. Obstet. Gynecol. 166: 1295–1299.[Medline]

70. Taufield, P. A., Ales, K. L., Resnick, L. M., Druzin, M. L., Gertner, J. M. & Laragh, J. H. (1987) Hypocalciuria in preeclampsia. N. Engl. J. Med. 316: 715–718.[Abstract]

71. Hubel, C. A., Roberts, J. M., Taylor, R. N., Musci, T. J., Rogers, G. M. & McLaughlin, M. K. (1989) Lipid peroxidation in pregnancy: new perspectives on preeclampsia. Am. J. Obstet. Gynecol. 161: 1025–1034.[Medline]

72. Wang, Y. P., Walsh, S. W., Guo, J. D. & Zhang, J. Y. (1991) The imbalance between thromboxane and prostacyclin in preeclampsia is associated with an imbalance between lipid peroxides and vitamin E in maternal blood. Am. J. Obstet. Gynecol. 165: 1695–1700.[Medline]

73. Schiff, E., Friedman, S. A., Stampfer, M., Kao, L., Barrett, P. H. & Sibai, B. M. (1996) Dietary consumption and plasma concentrations of vitamin E in pregnancies complicated by preeclampsia. Am. J. Obstet. Gynecol. 175: 1024–1028.[Medline]

74. Powers, R. W., Evans, R. W., Majors, A. K., Ojimba, J. I., Ness, R. B., Crombleholme, W. R. & Roberts, J. M. (1998) Plasma homocysteine concentration is increased in preeclampsia and is associated with evidence of endothelial activation. Am. J. Obstet. Gynecol. 179: 1605–1611.[Medline]

75. Langman, L. J. & Cole, D. E. (1999) Homocysteine. Crit. Rev. Clin. Lab. Sci. 36: 365–406.[Medline]

76. Selhub, J., Jacques, P. F., Wilson, P. W., Rush, D. & Rosemberg, I. H. (1993) Vitamin status and intake as primary determinants of homocysteinemia in an elderly population. JAMA 270: 2693–2698.[Abstract/Free Full Text]

77. Hankey, G. J. & Eikelboom, J. W. (1999) Homocysteine and vascular disease. Lancet 354: 407–413.[Medline]

78. Hubel, C. A. & Roberts, J. (1999) Lipid metabolism and oxidative stress. In: Chelsey's Hypertensive Disorders in Pregnancy (Lindheimer, M. D., Roberts, J. M. & Cunningham, F. G. eds.), pp. 453–486. Appleton and Lange, Stamford, CT.

79. Halliwell, B. (1995) Oxidation of low-density lipoproteins: questions of initiation, propagation, and the effect of antioxidants. Am. J. Clin. Nutr. 61: 670S–677S.[Abstract/Free Full Text]

80. Uotila, J., Solakivi, T., Jaakkola, O., Tuimala, R. & Lehtimaki, T. (1998) Antibodies against copper-oxidised and malondialdehyde-modified low density lipoproteins in pre-eclampsia pregnancies. Br. J. Obstet. Gynaecol. 105: 1113–1117.[Medline]

81. Kramer, M. S. (2002) Nutritional advice in pregnancy. Cochrane Database Syst. Rev. CD000149.

82. Kramer, M. S. (2002) Balanced protein/energy supplementation in pregnancy. Cochrane Database Syst. Rev. CD000032.

83. Kramer, M. S. (2002) Isocaloric balanced protein supplementation in pregnancy. Cochrane Database Syst. Rev. CD000118.

84. Kramer, M. S. (2002) Energy/protein restriction for high weight-for-height or weight gain during pregnancy. Cochrane Database Syst. Rev. CD000080.

85. Duley, L. & Henderson-Smart, D. (2002) Reduced salt intake compared to normal dietary salt, or high intake, in pregnancy. Cochrane Database Syst. Rev. CD001687.

86. Hofmeyr, G. J., Atallah, A. N. & Duley, L. (2002) Calcium supplementation during pregnancy for preventing hypertensive disorders and related problems (Cochrane Review). In: The Cochrane Library, Issue 4: CD 001059. Update Software, Oxford.

87. Makrides, M. & Crowther, C. A. (2002) Magnesium supplementation in pregnancy. Cochrane Database Syst. Rev. CD000937.

88. Duley, L. (1995) Prophylactic fish oil in pregnancy. In: Pregnancy and Childbirth Module of the Cochrane Database of Systematic Reviews (Enkin, M. W., Keirse, M. J., Renfrew, M. J. & Neilson, J. P., eds.). British Medical Journal Publishing Group, London, UK.

89. Mahomed, K. (2002) Iron and folate supplementation in pregnancy. Cochrane Database Syst. Rev. CD001135.

90. Mahomed, K. (2002) Folate supplementation in pregnancy. Cochrane Database Syst. Rev. CD000183.

91. Mahomed, K. (2002) Zinc supplementation in pregnancy. Cochrane Database Syst. Rev. CD000230.

92. Mahomed, K. (2002) Iron supplementation in pregnancy. Cochrane Database Syst. Rev. CD000117.

93. Cuervo, L. G. & Mahomed, K. (2002) Treatments for iron deficiency anaemia in pregnancy. Cochrane Database Syst. Rev. CD003094.

94. Conde-Agudelo, A. & Belizan, J. M. (2000) Risk factors for pre-eclampsia in a large cohort of Latin American and Caribbean women. Br. J. Obstet. Gynaecol. 107: 75–83.

95. Sibai, B. M., Ewell, M., Levine, R. J., Klebanoff, M. A., Esterlitz, J., Catalano, P. M., Goldenberg, R. L. & Joffe, G. (1997) Risk factors associated with preeclampsia in healthy nulliparous women. The Calcium for Preeclampsia Prevention (CPEP) Study Group. Am. J. Obstet. Gynecol. 177: 1003–1010.[Medline]

96. Eskenazi, B., Fenster, L. & Sidney, S. (1991) A multivariate analysis of risk factors for preeclampsia. JAMA 266: 237–241.[Abstract/Free Full Text]

97. Moutquin, J. M., Garner, P. R., Burrows, R. F., Rey, E., Helewa, M. E., Lange, I. R. & Rabkin, S. W. (1997) Report of the Canadian Hypertension Society Consensus Conference: 2. Nonpharmacologic management and prevention of hypertensive disorders in pregnancy. Can. Med. Assoc. J. 157: 907–919.[Abstract]

98. Dürr, J. & Lindheimer, M. (1998) Control of volume and body tonicity. in: chesley's hypertensive disorders in pregnancy (Lindheimer, R. J. C. F., ed.), pp. 103–166. Appleton & Lange, Stamford, CT.

99. Robinson, M. (1958) Salt in pregnancy. Lancet 1: 178–181.[Medline]

100. Bower, D. (1961) The influence of dietary salt intake on preeclampsia. J. Obstet. Gynaecol. Br. Commonw. 63: 123–126.

101. van der Maten, G. D. (1995) Low sodium diet in pregnancy: effects on maternal nutritional status. Eur. J. Obstet. Gynecol. Reprod. Biol. 61: 63–64.[Medline]

102. Villar, J. & Belizan, J. M. (2000) Same nutrient, different hypotheses: disparities in trials of calcium supplementation during pregnancy. Am. J. Clin. Nutr. 71: 1375S–1379S.[Abstract/Free Full Text]

103. Fleet, J. C. & Castillo, P. (2001) Magnesium. In: Present Knowledge in Nutrition (Bowman, B. A. & Russell, R. M., eds.), pp. 292–301. ILSI Press, Washington, DC.

104. Institute of Medicine. (1991) Nutrition during pregnancy, p. 329. National Academy Press, Washington, D.C.

105. Conradt, A., Weidinger, H. & Algayer, H. (1984) On the role of magnesium in fetal hypotrophy, pregnancy induced hypertension and preeclampsia. Magnesium 6: 68–76.

106. Doyle, W., Crawford, M., Wynn, A. H. & Wynn, S. W. (1989) Maternal magnesium intake and pregnancy outcome. Magnes. Res. 2: 205–210.[Medline]

107. Morris, M. C., Sacks, F. & Rosner, B. (1993) Does fish oil lower blood pressure? A meta-analysis of controlled trials. Circulation 88: 523–533.[Abstract/Free Full Text]

108. Olsen, S. F., Hansen, H. S., Sorensen, T. I., Jensen, B., Secher, N. J., Sommer, S. & Knudsen, L. B. (1986) Intake of marine fat, rich in (n–3)–polyunsaturated fatty acids, may increase birth weight by prolonging gestation. Lancet 2: 367–369.[Medline]

109. Secher, N. J. & Olsen, S. F. (1990) Fish-oil and pre-eclampsia. Br. J. Obstet. Gynaecol. 97: 1077–1079.[Medline]

110. Sandstead, H. H. (1985) W.O. Atwater memorial lecture. Zinc: essentiality for brain development and function. Nutr. Rev. 43: 129–137.[Medline]

111. Vallee, B. L. & Falchuk, K. H. (1993) The biochemical basis of zinc physiology. Physiol. Rev. 73: 79–118.

112. Golub, M. S., Keen, C. L., Gershwin, M. E. & Hendrickx, A. G. (1995) Developmental zinc deficiency and behavior. J. Nutr. 125: 2263S–2271S.

113. Brown, K. H., Peerson, J. M. & Allen, L. H. (1998) Effect of zinc supplementation on children's growth: a meta-analysis of intervention trials. Bibl. Nutr. Dieta 54: 76–83.

114. Caulfield, L. E., Zavaleta, N., Shankar, A. H. & Merialdi, M. (1998) Potential contribution of maternal zinc supplementation during pregnancy to maternal and child survival. Am. J. Clin. Nutr. 68: 499S–508S.[Abstract]

115. Ness, R. B. & Roberts, J. M. (1996) Heterogeneous causes constituting the single syndrome of preeclampsia: a hypothesis and its implications. Am. J. Obstet. Gynecol. 175: 1365–1370.[Medline]

116. Gratacos, E., Casals, E., Deulofeu, R., Cararach, V., Alonso, P. L. & Fortuny, A. (1998) Lipid peroxide and vitamin E patterns in pregnant women with different types of hypertension in pregnancy. Am. J. Obstet. Gynecol. 178: 1072–1076.[Medline]

117. Ziari, S. A., Mireles, V. L., Cantu, C. G., Cervantes, M., III, Idrisa, A., Bobsom, D., Tsin, A. T. & Glew, R. H. (1996) Serum vitamin A, vitamin E, and beta-carotene levels in preeclamptic women in northern Nigeria. Am. J. Perinatol. 13: 287–291.[Medline]

118. Hubel, C. A., Kagan, V. E., Kisin, E. R., McLaughlin, M. K. & Roberts, J. M. (1997) Increased ascorbate radical formation and ascorbate depletion in plasma from women with preeclampsia: implications for oxidative stress. Free Radic. Biol. Med. 23: 597–609.[Medline]

119. Oostenbrug, G. S., Mensink, R. P., van Houwelingen, A. C., Al, M. D. & Hornstra, G. (1998) Pregnancy-induced hypertension: maternal and neonatal plasma lipid-soluble antioxidant levels and its relationship with fatty acid unsaturation. Eur. J. Clin. Nutr. 52: 754–759.[Medline]

120. Gulmezoglu, A. M., Hofmeyr, G. J. & Oosthuisen, M. M. (1997) Antioxidants in the treatment of severe pre-eclampsia: an explanatory randomised controlled trial. Br. J. Obstet. Gynaecol. 104: 689–696.[Medline]

121. Pandey, S., Srivastava, R., Pandey, S., Mukerjee, D., Khattri, S. & Shanker, K. (2000) Effect of vitamin "A" on free radical cascade in pregnancy induced hypertension. Boll. Chim. Farm. 139: 98–102.[Medline]

122. Sibai, B. M., Lindheimer, M., Hauth, J., Caritis, S., VanDorsten, P., Klebanoff, M., MacPherson, C., Landon, M., Miodovnik, M., Paul, R., Meis, P. & Dombrowski, M. (1998) Risk factors for preeclampsia, abruptio placentae, and adverse neonatal outcomes among women with chronic hypertension. N. Engl. J. Med. 339: 667–671.[Abstract/Free Full Text]

123. Barton, J. R., O'Brien, J. M., Bergauer, N. K., Jacques, D. L. & Sibai, B. M. (2001) Mild gestational hypertension remote from term: progression and outcome. Am. J. Obstet. Gynecol. 184: 979–983.[Medline]

124. Ferrer, R. L., Sibai, B. M., Mulrow, C. D., Chiquette, E., Stevens, K. R. & Cornell, J. (2000) Management of mild chronic hypertension during pregnancy: a review. Obstet. Gynecol. 96: 849–860.[Medline]

125. Lindheimer, M. D. & Abalos, E. (2002) Management of high blood pressure in pregnancy. In: Calcium antagonists in clinical medicine (Epstein, M., ed.), pp. 507–534. Hanley and Belfus, Philadelphia.

126. Cook, J. D., Dassenko, S. A. & Whittaker, P. (1991) Calcium supplementation: effect on iron absorption. Am. J. Clin. Nutr. 53: 106–111.[Abstract/Free Full Text]

127. Gleerup, A., Rossander-Hulten, L. & Hallberg, L. (1993) Duration of the inhibitory effect of calcium on non-haem iron absorption in man. Eur. J. Clin. Nutr. 47: 875–879.[Medline]

128. Minihane, A. M. & Fairweather-Tait, S. J. (1998) Effect of calcium supplementation on daily nonheme-iron absorption and long-term iron status. Am. J. Clin. Nutr. 68: 96–102.[Abstract]

129. Ridwan, E., Schultink, W., Dillon, D. & Gross, R. (1996) Effects of weekly iron supplementation on pregnant Indonesian women are similar to those of daily supplementation. Am. J. Clin. Nutr. 63: 884–890.[Abstract/Free Full Text]

130. Mumtaz, Z., Shahab, S., Butt, N., Rab, M. A. & DeMuynck, A. (2000) Daily iron supplementation is more effective than twice weekly iron supplementation in pregnant women in Pakistan in a randomized double-blind clinical trial. J. Nutr. 130: 2697–2702.[Abstract/Free Full Text]

131. Young, M. W., Lupafya, E., Kapenda, E. & Bobrow, E. A. (2000) The effectiveness of weekly iron supplementation in pregnant women of rural northern Malawi. Trop. Doct. 30: 84–88.[Medline]

132. Mahale, A. R. & Shah, S. H. (1993) New schedule of intramuscular iron administration for pregnant women. Asia Oceania J. Obstet. Gynaecol. 19: 141–144.

133. Beaton, G. & McCabe, G. P. (2000) Efficacy of intermittent iron supplementation in the control of iron deficiency anemia in developing countries: analysis of experience to date. Micronutrient Initiative, Ottawa, Canada.

134. Makrides, M., Duley, L. & Olsen, S. F. (2002) Fish oil and other prostaglandin precursor supplementation during pregnancy for reducing pre-eclampsia, preterm birth, low birth weight and intrauterine growth restriction (Protocol for a Cochrane Review). The Cochrane Library, Issue 2. Update Software, Oxford. (in press)

135. Zavaleta, N., Caulfield, L. E. & Garcia, T. (2000) Changes in iron status during pregnancy in Peruvian women receiving prenatal iron and folic acid supplements with or without zinc. Am. J. Clin. Nutr. 71: 956–961.[Abstract/Free Full Text]

136. Kulier, R., Gulmezoglu, A. M., de Onis, M. & Villar, J. (2002) Vitamin A supplementation during pregnancy (Protocol for a Cochrane Review). The Cochrane Library, Issue 2. Update Software, Oxford. (in press)

137. Thomson, A. M. (1959) Maternal stature and reproductive efficiency. Eugen. Rev. 51: 157–162.

138. Bernard, R. (1952) The shape and size of the female pelvis. Edinburgh Med. J. 59: 1–15.

139. James, D. K. & Chiswick, M. L. (1979) Kielland's forceps: role of antenatal factors in prediction of use. BMJ 1: 10–11.

140. Dougherty, C. R. & Jones, A. D. (1988) Obstetric management and outcome related to maternal characteristics. Am. J. Obstet. Gynecol. 158: 470–474.[Medline]

141. Thomson, M. & Hanley, J. (1988) Factors predisposing to difficult labor in primiparas. Am. J. Obstet. Gynecol. 158: 1074–1078.[Medline]

142. Parsons, M. T., Winegar, A., Siefert, L. & Spellacy, W. N. (1989) Pregnancy outcomes in short women. J. Reprod. Med. 34: 357–361.[Medline]

143. Tsu, V. D. (1992) Maternal height and age: risk factors for cephalopelvic disproportion in Zimbabwe. Int. J. Epidemiol. 21: 941–946.[Abstract/Free Full Text]

144. Adadevoh, S. W., Hobbs, C. & Elkins, T. E. (1989) The relation of the true conjugate to maternal height and obstetric performance in Ghanaians. Int. J. Gynaecol. Obstet. 28: 243–251.[Medline]

145. Harrison, K. A. (1990) Predicting trends in operative delivery for cephalopelvic disproportion in Africa. Lancet 335: 861–862.

146. Merchant, K. M., Villar, J. & Kestler, E. (2001) Maternal height and newborn size relative to risk of intrapartum caesarean delivery and perinatal distress. Br. J. Obstet. Gynaecol. 108: 689–696.

147. Garner, P., Kramer, M. S. & Chalmers, I. (1992) Might efforts to increase birth weight in undernourished women do more harm than good? Lancet 340: 1021–1023.[Medline]

148. Vickery, C. (1993) Maternal supplementation and birth weight. Lancet 341: 53.[Medline]

149. de Onis, M., Villar, J. & Gulmezoglu, M. (1998) Nutritional interventions to prevent intrauterine growth retardation: evidence from randomized controlled trials. Eur. J. Clin. Nutr. 52 (Suppl. 1), S83–S93.

150. Gulmezoglu, M., de Onis, M. & Villar, J. (1997) Effectiveness of interventions to prevent or treat impaired fetal growth. Obstet. Gynecol. Surv. 52: 139–149.[Medline]

151. Atton, C. & Watney, P. G. M. (1990) Selective supplementation in pregnancy: effect on birth weight. J. Hum. Nutr. Diet. 3: 381–392.

152. McDonald, E. C., Pollitt, E., Mueller, W., Hsueh, A. M. & Sherwin, R. (1981) The Bacon Chow study: maternal nutrition supplementation and birth weight of offspring. Am. J. Clin. Nutr. 34: 2133–2144.[Abstract/Free Full Text]

153. Kardjati, S., Kusin, J. A. & De With, C. (1988) Energy supplementation in the last trimester of pregnancy in East Java: I. Effect on birth weight. Br. J. Obstet. Gynaecol. 95: 783–794.[Medline]

154. Viegas, O. A., Scott, P. H., Cole, T. J., Eaton, P., Needham, P. G. & Wharton, B. A. (1982) Dietary protein energy supplementation of pregnant Asian mothers at Sorrento, Birmingham. II: Selective during third trimester only. Br. Med. J. (Clin. Res. Ed.) 285: 592–595.

155. Viegas, O. A., Scott, P. H., Cole, T. J., Mansfield, H. N., Wharton, P. & Wharton, B. A. (1982) Dietary protein energy supplementation of pregnant Asian mothers at Sorrento, Birmingham. I: Unselective during second and third trimesters. Br. Med. J. (Clin. Res. Ed.) 285: 589–592.

156. Ceesay, S. M., Prentice, A. M., Cole, T. J., Foord, F., Weaver, L. T., Poskitt, E. M. & Whitehead, R. G. (1997) Effects on birth weight and perinatal mortality of maternal dietary supplements in rural Gambia: 5 year randomised controlled trial. BMJ 315: 786–790.[Abstract/Free Full Text]

157. Merchant, K., Martorell, R. & Haas, J. (1990) Maternal and fetal responses to the stresses of lactation concurrent with pregnancy and of short recuperative intervals. Am. J. Clin. Nutr. 52: 280–288.[Abstract/Free Full Text]

158. Rivera, J. A., Martorell, R., Ruel, M. T., Habicht, J. P. & Haas, J. D. (1995) Nutritional supplementation during the preschool years influences body size and composition of Guatemalan adolescents. J. Nutr. 125 (Suppl.), 1068S–1077S.

159. Martorell, R., Rivera, J. A., Schroeder, D. G., Ramakrishnan, U., Pollitt, E. & Ruel, M. T. (1995) Long term consequences of growth retardation during early childhood. Arch. Latinoam. Nutr. 45: 1095–1135.

160. Kramer, M. S. (2002) High protein supplementation in pregnancy. Cochrane Database Syst. Rev. CD000105.

161. Bucher, H. C., Guyatt, G. H., Cook, R. J., Hatala, R., Cook, D. J., Lang, J. D. & Hunt, D. (1996) Effect of calcium supplementation on pregnancy-induced hypertension and preeclampsia: a meta-analysis of randomized controlled trials. JAMA 275: 1113–1117.[Abstract/Free Full Text]

162. Boggess, K. A., Samuel, L., Schmucker, B. C., Waters, J. & Easterling, T. R. (1997) A randomized controlled trial of the effect of third-trimester calcium supplementation on maternal hemodynamic function. Obstet. Gynecol. 90: 157–161.[Medline]

163. Czeizel, A. E., Dudas, I. & Metneki, J. (1994) Pregnancy outcomes in a randomised controlled trial of periconceptional multivitamin supplementation. Final report. Arch. Gynecol. Obstet. 255: 131–139.[Medline]

164. Lumley, J., Watson, L., Watson, M. & Bower, C. (2002) Periconceptional supplementation with folate and/or multivitamins for preventing neural tube defects. Retraction in Lumley, J., Watson, L., Watson, M., Bower, C. Cochrane Database Syst. Rev. CD001056.

165. Magee, L. A., Mazzota, P. & Koren, G. (2002) Evidence-based view of safety and effectiveness of pharmacologic therapy for nausea and vomiting of pregnancy (NVP). Am. J. Obstet. Gynecol. 186 (Suppl.), S256–S261.

166. Merialdi, M., Caulfield, L. E., Zavaleta, N., Figueroa, A. & DiPietro, J. A. (1999) Adding zinc to prenatal iron and folate tablets improves fetal neurobehavioral development. Am. J. Obstet. Gynecol. 180: 483–490.[Medline]

167. World Health Organization. The WHO Reproductive Health Library. 5. 2002. WHO, Geneva.

168. Gulmezoglu, M., Villar, J., Hofmeyr, J., Duley, L. & Belizan, J. M. (1998) Randomised trials in maternal and perinatal medicine: global partnerships are the way forward. Br. J. Obstet. Gynaecol. 105: 1244–1247.[Medline]

169. Department of Reproductive Health and Research. World Health Organization and the Global Preeclampsia Eclampsia Collaboration. Global Programme to Conquer Preeclampsia/Eclampsia. 2002. WHO, Geneva.

170. Baird, D. (1985) Changing problems and priorities in obstetrics. Br. J. Obstet. Gynaecol. 92: 115–121.[Medline]

This article has been cited by other articles:

				G. Mwabu The Production of Child Health in Kenya: A Structural Model of Birth Weight J. Afr. Econ., March 1, 2009; 18(2): 212 - 260. [Abstract] [Full Text] [PDF]

				S. L Barber and P. J Gertler Empowering women to obtain high quality care: evidence from an evaluation of Mexico's conditional cash transfer programme Health Policy Plan., January 1, 2009; 24(1): 18 - 25. [Abstract] [Full Text] [PDF]

				G. Hiremath Micronutrient supplementation in pregnancy in developing countries BMJ, November 7, 2008; 337(nov07_4): a1942 - a1942. [Full Text]

				R. S. Gibson, Y. Abebe, S. Stabler, R. H. Allen, J. E. Westcott, B. J. Stoecker, N. F. Krebs, and K. M. Hambidge Zinc, Gravida, Infection, and Iron, but Not Vitamin B-12 or Folate Status, Predict Hemoglobin during Pregnancy in Southern Ethiopia J. Nutr., March 1, 2008; 138(3): 581 - 586. [Abstract] [Full Text] [PDF]

				A. T. Merchant, G. Msamanga, E. Villamor, E. Saathoff, M. O'Brien, E. Hertzmark, D. J. Hunter, and W. W. Fawzi Multivitamin Supplementation of HIV-Positive Women during Pregnancy Reduces Hypertension J. Nutr., July 1, 2005; 135(7): 1776 - 1781. [Abstract] [Full Text] [PDF]

				A. Vahratian, A. M. Siega-Riz, D. A. Savitz, and J. M. Thorp Jr. Multivitamin Use and the Risk of Preterm Birth Am. J. Epidemiol., November 1, 2004; 160(9): 886 - 892. [Abstract] [Full Text] [PDF]

				R. Baltussen, C. Knai, and M. Sharan Iron Fortification and Iron Supplementation are Cost-Effective Interventions to Reduce Iron Deficiency in Four Subregions of the World J. Nutr., October 1, 2004; 134(10): 2678 - 2684. [Abstract] [Full Text] [PDF]

				A. A. Jackson, Z. A. Bhutta, and P. Lumbiganon Introduction J. Nutr., May 1, 2003; 133(5): 1589S - 1591. [Full Text] [PDF]

				M. Merialdi, G. Carroli, J. Villar, E. Abalos, A. M. Gulmezoglu, R. Kulier, and M. de Onis Nutritional Interventions during Pregnancy for the Prevention or Treatment of Impaired Fetal Growth: An Overview of Randomized Controlled Trials J. Nutr., May 1, 2003; 133(5): 1626S - 1631. [Abstract] [Full Text] [PDF]

				J. Villar, M. Merialdi, A. M. Gulmezoglu, E. Abalos, G. Carroli, R. Kulier, and M. de Onis Characteristics of Randomized Controlled Trials Included in Systematic Reviews of Nutritional Interventions Reporting Maternal Morbidity, Mortality, Preterm Delivery, Intrauterine Growth Restriction and Small for Gestational Age and Birth Weight Outcomes J. Nutr., May 1, 2003; 133(5): 1632S - 1639. [Full Text] [PDF]

				C. H.D. Fall, C. S. Yajnik, S. Rao, A. A. Davies, N. Brown, and H. J.W. Farrant Micronutrients and Fetal Growth J. Nutr., May 1, 2003; 133(5): 1747S - 1756. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Villar, J. Articles by de Onis, M. Search for Related Content PubMed PubMed Citation Articles by Villar, J. Articles by de Onis, M.