Vitamin supplementation for preventing miscarriage

  • Review
  • Intervention

Authors

  • Alice Rumbold,

    Corresponding author
    1. The University of Adelaide, The Robinson Institute, Adelaide, NT, Australia
    • Alice Rumbold, The Robinson Institute, The University of Adelaide, Ground Floor, Norwich Centre, 55 King William Road, Adelaide, NT, SA 5006, Australia. alice.rumbold@adelaide.edu.au.

    Search for more papers by this author
  • Philippa Middleton,

    1. The University of Adelaide, ARCH: Australian Research Centre for Health of Women and Babies, Discipline of Obstetrics and Gynaecology, Adelaide, South Australia, Australia
    Search for more papers by this author
  • Ning Pan,

    1. The University of Adelaide, ARCH: Australian Research Centre for Health of Women and Babies, Discipline of Obstetrics and Gynaecology, Adelaide, South Australia, Australia
    Search for more papers by this author
  • Caroline A Crowther

    1. The University of Adelaide, ARCH: Australian Research Centre for Health of Women and Babies, Discipline of Obstetrics and Gynaecology, Adelaide, South Australia, Australia
    Search for more papers by this author

Abstract

Background

Miscarriage is a common complication of pregnancy that can be caused by a wide range of factors. Poor dietary intake of vitamins has been associated with an increased risk of miscarriage, therefore supplementing women with vitamins either prior to or in early pregnancy may help prevent miscarriage.

Objectives

The objectives of this review are to determine the effectiveness and safety of any vitamin supplementation, on the risk of spontaneous miscarriage, maternal adverse outcomes and fetal and infant adverse outcomes.

Search methods

We searched the Cochrane Pregnancy and Childbirth Group Trials Register (21 June 2010).

Selection criteria

All randomised and quasi-randomised trials comparing one or more vitamins with either placebo, other vitamins, no vitamins or other interventions, prior to conception, periconceptionally or in early pregnancy (less than 20 weeks' gestation).

Data collection and analysis

At least two review authors independently assessed trials for inclusion, extracted data and assessed trial quality.

Main results

We identified 28 trials assessing supplementation with any vitamin(s) starting prior to 20 weeks' gestation and reporting at least one primary outcome that was eligible for the review. Overall, the included trials involved 96,674 women and 98,267 pregnancies. Three trials were cluster randomised and combined contributed data for 62,669 women and 64,210 pregnancies in total. No significant differences were seen between women taking any vitamins compared with controls for total fetal loss (relative risk (RR) 1.04, 95% confidence interval (CI) 0.95 to 1.14), early or late miscarriage (RR 1.09, 95% CI 0.95 to 1.25) or stillbirth (RR 0.86, 95% CI 0.65 to 1.13) and most of the other primary outcomes, using fixed-effect models. Compared with controls, women given any type of vitamin(s) pre or peri-conception were more likely to have a multiple pregnancy (RR 1.38, 95% CI 1.12 to 1.70, three trials, 20,986 women).

Authors' conclusions

Taking any vitamin supplements prior to pregnancy or in early pregnancy does not prevent women experiencing miscarriage or stillbirth. However, women taking vitamin supplements may be more likely to have a multiple pregnancy. There is insufficient evidence to examine the effects of different combinations of vitamins on miscarriage, stillbirth or other maternal and infant outcomes.

Résumé scientifique

Supplémentation en vitamines pour prévenir les fausses couches

Contexte

La fausse couche est une complication fréquente de la grossesse qui peut être due à un large éventail de facteurs. Un apport alimentaire insuffisant en vitamines a été associé à un risque accru de fausses couches ; c’est pour quoi la supplémentation vitaminique des femmes avant une grossesse ou en début de grossesse peut contribuer à prévenir les fausses couches.

Objectifs

Cette revue a pour objectif de déterminer l'efficacité et l'innocuité de toute supplémentation vitaminique, sur le risque de fausse couche spontanée, les effets indésirables pour la mère et les conséquences indésirables pour le fœtus et le nourrisson.

Stratégie de recherche documentaire

Nous avons effectué des recherches dans le registre des essais cliniques du groupe Cochrane sur la grossesse et la naissance (21 juin 2010).

Critères de sélection

Tous les essais randomisés et quasi-randomisés comparant une ou plusieurs vitamines à un placebo, à d'autres vitamines, à aucun apport de vitamines ou à d'autres interventions, avant la conception, aux alentours de la conception ou en début de grossesse (moins de 20 semaines de gestation).

Recueil et analyse des données

Au moins deux auteurs ont évalué, de manière indépendante, les essais à inclure dans la revue, extrait les données et évalué la qualité des essais.

Résultats principaux

Nous avons identifié 28 essais évaluant la supplémentation avec tous types de vitamines démarrée avant 20 semaines de gestation et rapportant au moins un critère principal qui était éligible pour la revue. Dans l'ensemble, les essais inclus portaient sur 96 674 femmes et 98 267 grossesses. Trois essais ont été randomisés en cluster et une fois combinés, ont fourni des données sur 62 669 femmes et 64 210 grossesses au total. Aucune différence significative n'a été observée chez les femmes prenant des vitamines comparativement aux témoins en termes de perte fœtale totale (risque relatif (RR) = 1,04, intervalle de confiance (IC) à 95 % 0,95 à 1,14), fausses couches précoces ou tardives (RR = 1,09, IC à 95 % 0,95 à 1,25) ou mortinaissance (RR = 0,86, IC à 95 % 0,65 à 1,13) et pour la plupart des autres critères principaux, en utilisant des modèles à effets fixes. Comparativement aux témoins, les femmes ayant reçu tout type de vitamine(s) avant la conception ou aux alentours de la conception étaient davantage susceptibles d'avoir une grossesse multiple (RR = 1,38, IC à 95 % 1,12 à 1,70, trois essais, 20 986 femmes).

Conclusions des auteurs

La prise de suppléments vitaminiques avant la grossesse ou en début de grossesse n'empêche ni les fausses couches ni les mortinaissances. Cependant, les femmes qui prennent des suppléments vitaminiques sont davantage susceptibles d'avoir une grossesse multiple. Il n'y a pas de preuves suffisantes examinant les effets des différentes associations de vitamines sur les fausses couches, les mortinaissances ou d'autres résultats pour la mère et le nourrisson.

Resumo

Suplementos de vitaminas para prevenção do aborto

Introdução

O aborto é uma complicação comum da gravidez que pode ser causada por muitos fatores. Uma dieta pobre em vitaminas tem sido associada a um aumento no risco de aborto. Portanto, dar suplementos vitamínicos antes da gravidez ou no seu início poderia ajudar a prevenir o aborto.

Objetivos

Os objetivos desta revisão são avaliar a efetividade e a segurança de qualquer suplementação vitamínica para prevenção do abortamento espontâneo, desfechos adversos maternos, fetais e infantis.

Métodos de busca

Pesquisamos na base de dados Cochrane Pregnancy and Childbirth Group Trials Register (até 21 de junho 2010).

Critério de seleção

Selecionamos todos ensaios clínicos randomizados ou quasi-randomizados que compararam uma ou mais vitaminas versus placebo, outras vitaminas, nenhuma vitamina ou outras intervenções, antes de engravidar, no período periconcepcional ou no início da gestação (antes da 20ª semana).

Coleta dos dados e análises

Pelo menos dois autores selecionaram independentemente os estudos, extraíram os dados e avaliaram a qualidade dos ensaios clínicos.

Principais resultados

Identificamos 28 estudos que avaliaram a suplementação vitamínica com qualquer vitamina(s), iniciada antes da 20ª semana de gestação, e que relatavam pelo menos um dos desfechos primários desta revisão. No total, os estudos incluídos envolveram 96.674 mulheres e 98.267 gestações. Três estudos eram do tipo cluster (randomização por conglomerados) e totalizavam 62.669 mulheres e 64.210 gravidezes. Não houve diferença significativa entre as mulheres que tomaram qualquer vitamina quando comparadas com os controles para perda fetal em geral (risco relativo, RR, 1,04, intervalo de confiança de 95%, 95% CI, de 0,95 a 1,14), aborto precoce ou tardio (RR 1,09, 95% CI 0,95 a 1,25) ou natimortos (RR 0,86, 95% CI 0,65 a 1,13) e a maioria de outros desfechos primários. Essas metanálises foram feitas usando modelos de efeito fixo. As mulheres que receberam qualquer tipo de vitamina(s) antes de engravidar ou no período peri-concepcional tiveram maior probabilidade de ter uma gravidez múltipla (RR 1,38, 95% CI 1,12 a 1,70, três estudos com 20.986 mulheres) do que as mulheres no grupo controle.

Conclusão dos autores

Tomar qualquer tipo de suplemento vitamínico antes ou no início da gravidez não previne o aborto ou que o feto seja natimorto. Entretanto, mulheres que fazem suplementação vitamínica têm maior chance de ter gravidez múltipla. Não há evidência suficiente para se avaliar o efeito de diferentes combinações de vitaminas sobre o risco de aborto, a taxa de natimortos ou outros desfechos maternos ou infantis.

Notas de tradução

Tradução do Centro Cochrane do Brasil (Telma Lima).

Plain language summary

Vitamin supplementation for preventing miscarriage

Supplementing women with any vitamins does not reduce the number of women who miscarry or have a stillbirth.

Poor diet, without enough vitamins, has been associated with an increased risk of women losing their baby in early pregnancy. Taking vitamin supplements prior to pregnancy or in early pregnancy may reduce the risk of miscarriage, but this review did not find this to be the case. However, women taking vitamin supplements before or at the time of conception may be more likely to have a multiple pregnancy. More research is needed to determine the impact of different combinations of vitamins. This review included 28 trials involving 96,674 women (98,267 pregnancies).

Résumé simplifié

Supplémentation en vitamines pour prévenir les fausses couches

L'administration de suppléments vitaminiques aux femmes enceintes ne réduit pas le nombre de fausses couches ou de mortinaissances.

Une mauvaise alimentation, carencée en vitamines, a été associée chez la femme enceinte à un risque accru de fausse couche en début de grossesse. La prise de suppléments vitaminiques avant la grossesse ou en début de grossesse peut réduire le risque de fausse couche, mais cette revue n'a pas trouvé que c'était le cas. Cependant, les femmes qui prennent des suppléments vitaminiques avant ou au moment de la conception sont davantage susceptibles d’avoir une grossesse multiple. Des recherches supplémentaires sont nécessaires pour déterminer l'impact des différentes associations de vitamines. Cette revue comprenait 28 essais portant sur 96 674 femmes (98 267 grossesses).

Notes de traduction

Traduit par: French Cochrane Centre 12th November, 2012
Traduction financée par: Ministère du Travail, de l'Emploi et de la Santé Français

Resumo para leigos

Suplementação de vitaminas para prevenir o aborto

Tomar suplementos vitamínicos de qualquer tipo não reduz o número de mulheres que abortam ou têm um natimorto.

Dietas pobres, sem vitaminas suficientes, têm sido associadas a um aumento no risco de mulheres perderem seus bebes no início da gravidez. Tomar suplemento vitamínico antes ou no início da gravidez poderia reduzir o risco de aborto, mas esta revisão não confirma esse dado. No entanto, mulheres fazendo uso de suplementos vitamínicos antes ou no momento da concepção têm maior probabilidade de ter gravidez múltipla (gerar mais de um bebê). Mais pesquisas são necessárias para analisar o impacto de diferentes combinações de vitaminas. Esta revisão incluiu 28 estudos envolvendo 96.674 mulheres (96.267 gestações).

Notas de tradução

Tradução do Centro Cochrane do Brasil (Telma Lima).

Summary of findings(Explanation)

Summary of findings for the main comparison. Vitamin A versus placebo for preventing miscarriage
  1. 1 Follow-up up to 24 weeks post birth.
    2 Control group risk extracted from the original trial reference.
    3 Only one study for this outcome; therefore no point deducted for inconsistency between studies.
    4 Gestational age may have been underestimated, because women may have mistaken vaginal bleeding early in pregnancy for menses.

Vitamin A versus placebo for preventing miscarriage
Patient or population: pregnant women
Settings:
Intervention: vitamin A
Comparison: placebo
OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No of Participants
(studies)
Quality of the evidence
(GRADE)
Comments
Assumed riskCorresponding risk
placebovitamin A
Total fetal loss (including miscarriages or combined miscarriages and stillbirths)
Follow-up: 24 weeks1
83 per 1000 286 per 1000
(76 to 97)
RR 1.04
(0.92 to 1.17)
11723
(1 study)
⊕⊕⊕⊕
high 3
 
Neonatal death
Follow-up: 28 days
46 per 1000 250 per 1000
(42 to 60)
RR 1.09
(0.92 to 1.3)
10214
(1 study)
⊕⊕⊕⊕
high 3
 
Preterm birth282 per 1000 2293 per 1000
(251 to 341)
RR 1.04
(0.89 to 1.21)
11723
(1 study)
⊕⊕⊕⊝
moderate 3,4
 
*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio;
GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

Summary of findings 2 Vitamin A versus B-Carotene for preventing miscarriage

Summary of findings 2. Vitamin A versus B-Carotene for preventing miscarriage
  1. 1 Follow-up up to 24 weeks post birth.
    2 Control group risk extracted from the original trial reference.
    3 Only one study for this outcome; therefore no point deducted for inconsistency between studies.
    4 Gestational age may have been underestimated, because women may have mistaken vaginal bleeding early in pregnancy for menses.

Vitamin A versus B-Carotene for preventing miscarriage
Patient or population: pregnant women
Settings:
Intervention: vitamin A
Comparison: B-Carotene
OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No of Participants
(studies)
Quality of the evidence
(GRADE)
Comments
Assumed riskCorresponding risk
B-Carotenevitamin A
Total fetal loss (including miscarriages or combined miscarriages and stillbirths)
Follow-up: 24 weeks1
95 per 1000 296 per 1000
(86 to 108)
RR 1.01
(0.9 to 1.14)
11720
(1 study)
⊕⊕⊕⊕
high 3
 
Neonatal death
Follow-up: 28 days
50 per 1000 250 per 1000
(42 to 59)
RR 1
(0.85 to 1.18)
10228
(1 study)
⊕⊕⊕⊕
high 3
 
Preterm birth284 per 1000 2293 per 1000
(250 to 341)
RR 1.03
(0.88 to 1.2)
11720
(1 study)
⊕⊕⊕⊝
moderate 3,4
 
*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio;
GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

Summary of findings 3 Vitamin A plus iron plus folate versus iron plus folate for preventing miscarriage

Summary of findings 3. Vitamin A plus iron plus folate versus iron plus folate for preventing miscarriage
  1. 1 Control group risk extracted from the original trial reference(s).
    2 One study had incomplete data (of 243 pregnant women initially enrolled, 182 attended the postpartum examination; reasons for all dropouts not reported); however, this study contributed less than 30% of the weight to this analysis, and a point has not been deducted.
    3 One study included only women with HIV.
    4 One study included only women with anaemia.
    5 Wide 95% CIs.
    6 Only one study for this outcome; therefore no point deducted for inconsistency between studies.

Vitamin A plus iron plus folate versus iron plus folate for preventing miscarriage
Patient or population: pregnant women
Settings:
Intervention: vitamin A plus iron plus folate
Comparison: iron plus folate
OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No of Participants
(studies)
Quality of the evidence
(GRADE)
Comments
Assumed riskCorresponding risk
iron plus folatevitamin A plus iron plus folate
Total fetal loss (including miscarriages or combined miscarriages and stillbirths)Low risk populationRR 1.01
(0.61 to 1.66)
1640
(3 studies)
⊕⊝⊝⊝
very low 2,3,4,5
 
25 per 1000 125 per 1000
(15 to 41)
Medium risk population
31 per 1000 131 per 1000
(19 to 51)
High risk population
52 per 1000 152 per 1000
(32 to 86)
StillbirthStudy populationRR 1.29
(0.57 to 2.91)
1640
(3 studies)
⊕⊝⊝⊝
very low 2,3,4,5
 
14 per 100018 per 1000
(8 to 41)
Medium risk population
17 per 100022 per 1000
(10 to 49)
Preterm birth56 per 1000 162 per 1000
(33 to 117)
RR 1.11
(0.59 to 2.09)
700
(1 study)
⊕⊕⊝⊝
low 4,5,6
 
Birthweight
(grams)
The mean birthweight in the control groups was
2,805
The mean Birthweight in the intervention groups was
90 higher
(2.68 to 177.32 higher)
 594
(1 study)
⊕⊕⊝⊝
low 3,6
 
Maternal anaemiaStudy populationRR 0.96
(0.82 to 1.12)
700
(1 study)
⊕⊕⊕⊝
moderate 4,6
 
513 per 1000492 per 1000
(421 to 575)
Medium risk population
513 per 1000492 per 1000
(421 to 575)
*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio;
GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

Background

Miscarriage or pregnancy loss within the first 20 weeks of gestation is a frequent complication of pregnancy, with 12% to 16% of all clinically recognised pregnancies ending in miscarriage (Everett 1997; Regan 1989). Recurrent miscarriage, defined as the occurrence of three or more consecutive spontaneous miscarriages, affects one to two per cent of women of reproductive age (Coulam 1991). Miscarriage is associated with significant maternal morbidity including haemorrhage and infection and, in some cases, maternal death (NHMRC 2001), with maternal death more common in countries that are resource-poor (Goyaux 2001). Women experiencing miscarriage may suffer significant psychological and emotional stress. Grief experienced by women and their families can be complicated by feelings of self-blame, anxiety and depression, and social withdrawal and marital disturbances may result (Lee 1996). This emotional distress may be further compounded when women experience recurrent miscarriage.

Care of women suffering miscarriage is complicated by different definitions, disagreement on its validity as a clinical disorder and poor quality data on women not requiring hospital or outpatient treatment. Miscarriage can be caused by a wide range of factors and determining the aetiology of miscarriage is often difficult, with a variety of underlying mechanisms being potentially responsible. Up to two-thirds of early pregnancy losses (miscarriage before 12 weeks' gestation), are associated with chromosomal abnormalities (Stern 1996). While early miscarriages are more likely to be associated with chromosomal abnormalities and defective placental development, late miscarriage (miscarriage occurring between 12 and 20 weeks' gestation) may be more likely due to structural problems of the uterus and/or cervix, such as cervical incompetence. Women experiencing recurrent miscarriage may often have an underlying medical condition such as autoimmune disease, i.e. systemic lupus erythematosus and antiphospholipid syndrome, or other blood clotting disorders such as hyperhomocysteinemia (high levels of homocysteine in the blood) or another thrombophilia (Preston 1996). Recurrent miscarriage is thought to involve an underlying placental vascular pathology seen also in pre-eclampsia, intrauterine growth restriction and placental abruption (Ray 1999). Current therapies being considered for the prevention of miscarriage include progesterone and immunotherapy and these topics are covered in other Cochrane reviews (Haas 2009; Porter 2006).

Vitamins are essential nutrients required for a range of functions in the body. Vitamins are either water soluble, such as vitamin C and the B group vitamins (including folate), or fat soluble such as vitamin A, D, E and K. Vitamins are obtained from the diet and also from dietary supplements of either individual vitamin preparations or a multivitamin preparation. Multivitamins contain a range of vitamins and minerals, usually in doses similar to the recommended dietary intakes. Folate, vitamin B6 and vitamin B12 have been recommended for women with hyperhomocysteinemia, and therefore supplementation may influence the risk of spontaneous miscarriage in these women. Similarly, oxidative stress, where there is overproduction of reactive oxygen molecules leading to decreased levels of antioxidants, has been linked to spontaneous and recurrent miscarriage (Jauniaux 2000; Simsek 1998). Therefore, intake of antioxidant vitamins such as vitamin C and vitamin E may be an important factor associated with the risk of miscarriage. One observational study has demonstrated an association between the risk of spontaneous early miscarriage and dietary factors, with a high risk associated with poor intake of green vegetables, fruit and dairy products coupled with a high intake of fat (Di Cintio 2001). Little information is available about the impact of vitamins on the risk of early versus late miscarriage; however, dietary factors could theoretically influence structural abnormalities such as cervical incompetence. There is a growing body of research investigating the relationship between nutrition and placental development, fetal growth, pregnancy outcomes and adult diseases (Godfrey 1996; Morris 2001; NRC 1989). Therefore, adequate maternal nutrition, particularly vitamin intake, may be an important factor in preventing spontaneous miscarriage. There is currently little information about the most appropriate vitamin type or combination; therefore, this review will cover all vitamin types. Similarly, many commercially available vitamin preparations contain a range of combinations of vitamins; therefore, this review will pool all vitamin types together initially to determine overall if any vitamins and any improvement in maternal nutrition helps to prevent miscarriage. We will also discuss individual vitamin types separately.

The use of any vitamin supplements in pregnancy needs to be carefully monitored and evaluated for safety and efficacy. This is particularly true for early pregnancy use, where there is the potential for teratogenicity. High maternal levels of preformed vitamin A (retinoic acid) are known to induce spontaneous miscarriage and malformations involving the central nervous systems and cardiac development (WHO 1998). Potential teratogenic effects of other vitamins have not been reported. In contrast, folate supplementation and multivitamin supplementation have been associated with a decreased risk of neural tube defects (Lumley 2001b). Periconceptional folate and multivitamin supplementation for the prevention of neural tube defects has been covered in another Cochrane review (Lumley 2001b). Concerns have been raised about the impact of folate and multivitamin supplementation on the rate of multiple births (Lumley 2001a; Lumley 2001b); however, further research is needed to establish a direct causal relationship. Potential side effects of vitamins can occur with hypervitaminosis (excessive ingestion of one or more vitamins). Hypervitaminosis A (vitamin A poisoning) has been associated with irritability, fatigue, changes in the skin, hair loss, headache, and abdominal discomfort (Olsen 1999). Similarly hypervitaminosis D (vitamin D poisoning) has been associated with nausea, vomiting, weakness, disturbed digestion, and elevated blood and tissue calcium levels (Olsen 1999). For vitamin E, non-randomised controlled clinical trials of vitamin E supplementation in a variety of doses have failed to demonstrate any consistent side effects (Bendich 1993). However, there has been limited evaluation of the use of these and other vitamins in pregnancy. While water soluble vitamins such as vitamin C and the B group vitamins are easily excreted by the body, the fat soluble vitamins A, D, E and K may accumulate in the body and in the developing fetus. The safety of using these vitamins needs to be clearly demonstrated before they can be recommended for routine antenatal care.

The aims of this review are to identify all published and unpublished randomised controlled trials that investigate vitamins for the prevention of miscarriage and to assess the benefits and hazards of women using vitamins for the prevention of miscarriage.

Objectives

The objectives of this review are:

(1) to determine the effectiveness and safety of any vitamin supplementation taken by women prior to conception, periconceptionally and in early pregnancy on the risk of:

  • spontaneous miscarriage;

  • maternal adverse outcomes;

  • fetal and infant adverse outcomes.

(2) If vitamins are effective, to determine which of these agents are best and to compare vitamins with other interventions.

Methods

Criteria for considering studies for this review

Types of studies

All randomised trials and quasi-randomised trials comparing one or more vitamins with either placebo, other vitamins, no vitamins or other interventions, prior to conception, periconceptionally or in early to mid-pregnancy.

The review authors deemed it important to include any supplementation trials, where supplementation began prior to 20 weeks' gestation, and where at least one primary outcome as specified in the review was reported, even if the intervention was not specifically for the prevention of miscarriage. We excluded trials where the onset of supplementation occurred definitely after 20 weeks' gestation or where it was reported that the majority of women commenced supplementation after 20 weeks' gestation. We included trials where the onset of supplementation occurred both prior to and after 20 weeks' gestation, and when it could not be established whether the majority of the women started supplementation prior to 20 weeks' gestation. As these trials included some women over 20 weeks' gestation, who do not meet the eligibility criteria for the review, the main analyses for fetal loss outcomes were subgrouped according to the duration of vitamin usage, based on the categorisation of gestation at trial entry: before pregnancy, <12 weeks' gestation, between 12-20 weeks' gestation or 'mixed' which included women enrolled before and after 20 weeks' gestation).

Types of participants

Pregnant women (less than 20 weeks' gestation) or women in the reproductive age group planning on becoming pregnant in the near future, regardless of whether they are at low or high risk of having a miscarriage.

Types of interventions

Comparisons of any vitamin(s) alone or in combination with other agents with either placebo, other vitamin(s), no vitamin(s) or other interventions for the prevention of miscarriage, either in areas where there is inadequate dietary intake or where there is presumed adequate intake of that vitamin(s).

Types of outcome measures

Primary outcomes
For the woman
  1. Total fetal loss, defined as the combined numbers of early miscarriage (spontaneous pregnancy loss less than 12 weeks' gestation), late miscarriage (spontaneous pregnancy loss greater than or equal to 12 and less than 20 weeks), and stillbirth (pregnancy loss at greater than or equal to 20 weeks).

  2. Early or late miscarriage.

  3. Placental abruption.

  4. Pre-eclampsia.

  5. Psychological effects (anxiety and depression).

For the infant
  1. Stillbirth, perinatal or neonatal death.

  2. Preterm birth (defined as birth less than 37 weeks' gestation).

  3. Birthweight.

  4. Small-for-gestational age (birthweight less than the third centile or the most extreme centile reported).

  5. Congenital malformations.

To overcome wide variation in the definitions of miscarriage and stillbirth between studies, we included the combined outcome 'total fetal loss' in the review.

Secondary outcomes
  1. Multiple pregnancy (including only trials supplementing women prior to or around the time of conception).

  2. Very preterm birth (defined as less than 34 weeks' gestation).

  3. Apgar score less than seven at five minutes.

  4. Use of blood transfusion for the mother.

  5. Anaemia (maternal and infant).

  6. Placental weight.

  7. Methods of feeding: breastfeeding, formula or both.

  8. Subsequent fertility (subsequent pregnancy rate per couple or as defined by the authors).

  9. Poor growth at childhood follow-up.

  10. Disability at childhood follow-up.

  11. Adverse effects of vitamin supplementation sufficient to stop supplementation, such as manifestations of hypervitaminosis, headache, nausea, vomiting, diarrhoea.

  12. Maternal views of care.

Use of health service resources
  1. Gynaecological hospital admission.

  2. Admission to neonatal intensive care unit.

  3. Healthcare costs.

Search methods for identification of studies

Electronic searches

We searched the Cochrane Pregnancy and Childbirth Group Trials Register by contacting the Trials Search Co-ordinator (21 June 2010).

The Cochrane Pregnancy and Childbirth Group’s Trials Register is maintained by the Trials Search Co-ordinator and contains trials identified from: 

  1. quarterly searches of the Cochrane Central Register of Controlled Trials (CENTRAL);

  2. weekly searches of MEDLINE;

  3. handsearches of 30 journals and the proceedings of major conferences;

  4. weekly current awareness alerts for a further 44 journals plus monthly BioMed Central email alerts.

Details of the search strategies for CENTRAL and MEDLINE, the list of handsearched journals and conference proceedings, and the list of journals reviewed via the current awareness service can be found in the ‘Specialized Register’ section within the editorial information about the Cochrane Pregnancy and Childbirth Group

Trials identified through the searching activities described above are each assigned to a review topic (or topics). The Trials Search Co-ordinator searches the register for each review using the topic list rather than keywords. 

We did not apply any language restrictions.

For details of additional searching carried out for the previous version of the review, see Appendix 1.

Data collection and analysis

Selection of studies

Two review authors independently assessed for inclusion all the potential studies identified as a result of the search strategy. We resolved any disagreement through discussion.

Data extraction and management

We designed a form to extract data. For eligible studies, two review authors extracted the data using the agreed form. We resolved discrepancies through discussion. We entered data into Review Manager software (RevMan 2008) and checked for accuracy.

When information regarding any of the above was unclear, we attempted to contact authors of the original reports to provide further details.

Assessment of risk of bias in included studies

Two review authors independently assessed risk of bias for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2009). We resolved any disagreement by discussion or by involving a third review author.

(1) Sequence generation (checking for possible selection bias)

We describe for each included study the method used to generate the allocation sequence in sufficient detail to allow an assessment of whether it should produce comparable groups.

We assessed the method as:

  • adequate (any truly random process, e.g. random number table; computer random number generator),

  • inadequate (any non-random process, e.g. odd or even date of birth; hospital or clinic record number) or,

  • unclear.   

 (2) Allocation concealment (checking for possible selection bias)

We describe for each included study the method used to conceal the allocation sequence and determine whether intervention allocation could have been foreseen in advance of, or during recruitment, or changed after assignment.

We assessed the methods as:

  • adequate (e.g. telephone or central randomisation; consecutively numbered sealed opaque envelopes);

  • inadequate (open random allocation; unsealed or non-opaque envelopes, alternation; date of birth);

  • unclear.   

(3) Blinding (checking for possible performance bias)

We described for each included study the methods used, if any, to blind study participants and personnel from knowledge of which intervention a participant received. We will consider that studies are at low risk of bias if they were blinded, or if we judge that the lack of blinding could not have affected the results. We will assess blinding separately for different outcomes or classes of outcomes.

We assessed the methods as:

  • adequate, inadequate or unclear for participants;

  • adequate, inadequate or unclear for personnel;

  • adequate, inadequate or unclear for outcome assessors.

(4) Incomplete outcome data (checking for possible attrition bias through withdrawals, dropouts, protocol deviations)

We describe for each included study, and for each outcome or class of outcomes, the completeness of data including attrition and exclusions from the analysis. We state whether attrition and exclusions were reported, the numbers included in the analysis at each stage (compared with the total randomised participants), reasons for attrition or exclusion where reported, and whether missing data were balanced across groups or were related to outcomes. Where sufficient information was reported, or supplied by the trial authors, we re-included missing data in the analyses which we undertook. We assessed methods as:

  • adequate;

  • inadequate;

  • unclear.

(5) Selective reporting bias

We describe for each included study how we investigated the possibility of selective outcome reporting bias and what we found.

We assess the methods as:

  • adequate (where it is clear that all of the study’s pre-specified outcomes and all expected outcomes of interest to the review have been reported);

  • inadequate (where not all the study’s pre-specified outcomes have been reported; one or more reported primary outcomes were not pre-specified; outcomes of interest are reported incompletely and so cannot be used; study fails to include results of a key outcome that would have been expected to have been reported);

  • unclear.

Measures of treatment effect

Dichotomous data

For dichotomous data, we present results as summary risk ratio (RR) with 95% confidence intervals (CI). 

Continuous data

For continuous data, we used the mean difference if outcomes were measured in the same way between trials. We use the standardised mean difference to combine trials that measured the same outcome, but used different methods.  

Unit of analysis issues

Where trials recruited women prior to becoming pregnant, we reported the denominators for each trial as all women randomised; or where there was accurate information about the number of women in each trial who became pregnant, we reported the denominators as the number of women randomised and with a confirmed pregnancy.

We included all included trials in the initial analyses which we performed by any vitamin to include all vitamin combinations and then by individual vitamin type.

Cluster-randomised trials

We included cluster-randomised trials in the analyses along with individually randomised trials. We adopted a generic inverse variance approach for the meta-analyses for dichotomous outcomes where trials using cluster-randomisation techniques were included (Alderson 2004).

Dealing with missing data

For included studies, we noted levels of attrition. We explored the impact of including studies with high levels of missing data in the overall assessment of treatment effect by using Sensitivity analysis.

For all outcomes, we carried out analyses, as far as possible, on an intention-to-treat basis, i.e. we attempted to include all participants randomised to each group in the analyses, and analysed all participants in the group to which they were allocated, regardless of whether or not they received the allocated intervention. The denominator for each outcome in each trial was the number randomised minus any participants whose outcomes were known to be missing.

Assessment of heterogeneity

We applied tests of heterogeneity between trials to assess the significance of any differences between trials in the analyses (I2 greater than or equal to 30%) and we explored possible causes of heterogeneity.

Data synthesis

We carried out statistical analysis using the Review Manager software (RevMan 2008). We used fixed-effect meta-analysis for combining data where it is reasonable to assume that studies are estimating the same underlying treatment effect: i.e. where trials are examining the same intervention, and we judged the trials’ populations and methods sufficiently similar. If there is clinical heterogeneity sufficient to expect that the underlying treatment effects differ between trials, or if we detect substantial statistical heterogeneity, we used random-effects meta-analysis to produce an overall summary if an average treatment effect across trials is considered clinically meaningful. We treated the random-effects summary as the average range of possible treatment effects and we discuss the clinical implications of treatment effects differing between trials. If the average treatment effect is not clinically meaningful we would not combine trials.

Where we used random-effects analyses, we have presented the results as the average treatment effect with its 95% confidence interval.

Subgroup analysis and investigation of heterogeneity

Where included studies had more than two treatment arms or compared one or more vitamin intervention with another, we have not included these studies in the main analysis of 'any vitamins versus no or minimal vitamins'. We used data from these trials only in the subgroup analyses according to vitamin type.

As some of the trials included women over 20 weeks' gestation, who do not meet the eligibility criteria for the review, we have subgrouped the main analyses for fetal loss outcomes according to the duration of vitamin usage subgroup.

Where possible, we classified women into subgroups based on:

  1. the type of vitamin (vitamin C, vitamin A, multivitamin and folic acid);

  2. the duration of vitamin usage, based on time of trial entry: before pregnancy, < 12 weeks' gestation, between 12-20 weeks' gestation or 'mixed', which included women enrolled before and after 20 weeks' gestation;

  3. the dose of vitamin(s) (below or above the recommended dietary intake);

  4. their risk of spontaneous miscarriage (high risk defined as the presence of medical conditions associated with miscarriage such as hyperhomocysteinemia, thrombophilia, antiphospholipid syndrome, systemic lupus erythematosus; low risk defined as none of the above conditions); their risk of recurrent miscarriage (high risk defined as two or more previous consecutive spontaneous miscarriages, and/or the presence of medical conditions associated with miscarriage such as hyperhomocysteinemia, thrombophilia, antiphospholipid syndrome, systemic lupus erythematosus; low risk defined as none of the above conditions);

  5. low or adequate dietary vitamin intake at trial entry (low intake defined as less than the recommended daily intake for each vitamin in that setting, as measured by dietary questionnaire).

If we identified substantial heterogeneity, we performed subgroup analyses for the main fetal loss outcomes (total fetal loss, early or late miscarriage, stillbirth) by the above subgroups (where possible). We modelled heterogeneity that was not explained by subgroup analyses using a random-effects analysis.

For fixed-effect inverse variance meta-analyses we assessed differences between subgroups by interaction tests. For random-effects and fixed-effect meta-analyses using methods other than inverse variance, we assessed differences between subgroups by inspection of the subgroups’ confidence intervals; non-overlapping confidence intervals indicate a statistically significant difference in treatment effect between the subgroups.

Sensitivity analysis

We carried out sensitivity analyses to explore the effect of trial quality on the primary outcomes related to fetal loss (total fetal loss, early or late miscarriage, stillbirth). This involved restricting the analyses to trials with an 'adequate' rating of allocation concealment. We considered these trials high quality.

Results

Description of studies

See tables 'Characteristics of included studies' and 'Characteristics of excluded studies' for details of individual studies.

Included studies

We identified 28 trials assessing supplementation with any vitamin(s) starting prior to 20 weeks' gestation. The included trials involved 34,005 women plus a further 62,669 women who were enrolled in the three cluster randomised trials. Two of the trials (one cluster and one small trial) included women who were pregnant more than once in the study period, resulting in data being contributed for 64,210 pregnancies from the cluster trials, and 34,057 pregnancies for the individual trials. Many of the trials assessed interventions not specifically for the prevention of miscarriage. Four of the included studies were assessing folic acid supplementation for the prevention of neural tube defects (Czeizel 1994; ICMR 2000; Kirke 1992; MRC 1991). For the purpose of this review however, the authors deemed it important to include any supplementation trials, where supplementation began prior to 20 weeks' gestation, and where at least one main outcome as specified in the review was reported.

Participants

The demographic and obstetric characteristics of women varied widely between the trials (See table 'Characteristics of included studies'), as did the gestational age at trial entry and the type of supplements. Some trials enrolled women prior to conception (Christian 2003; Czeizel 1994; Hemmi 2003; ICMR 2000; Kirke 1992; MRC 1991) and asked women to continue taking the supplements up until the second or third missed menstrual period. One trial (Katz 2000) supplemented women from before to conception, through pregnancy and up to 3.5 years postpartum. Other trials enrolled women in the first trimester (Briscoe 1959; Rumiris 2006) or in early to mid pregnancy (Chappell 1999; Correia 1982; Fawzi 1998; Fawzi 2007; Fleming 1968; Fleming 1986; Kumwenda 2002; Osrin 2005; People's League 1942; Roberfroid 2008; Rumbold 2006; Rush 1980; Schmidt 2001; Steyn 2003; Taylor 1982; Van den Broek 2006; Villar 2009). Some of the trials enrolling women in early to mid pregnancy included women enrolled at or after 20 weeks' gestation (Chappell 1999; Fawzi 1998; Fawzi 2007; Fleming 1968; Fleming 1986; Kumwenda 2002; Osrin 2005; People's League 1942; Roberfroid 2008; Rumbold 2006; Rush 1980; Schmidt 2001; Spinnato 2007; Steyn 2003; Van den Broek 2006; Villar 2009); however, none of these studies separately reported data on the proportion of women who met the eligibility criteria for the review (i.e. less than 20 weeks' gestation). One trial (The Summit 2008), enrolled 41,839 women at 'any gestational age', however more than 70% of women were enrolled in the first or second trimester. Two trials (Fawzi 1998; Kumwenda 2002) involved vitamin A supplementation in women seropositive for the Human Immunodeficiency Virus (HIV).

The trials were conducted in both resource-rich and resource-poor countries including the United States (Briscoe 1959; Rush 1980), the United Kingdom (Chappell 1999; People's League 1942; Taylor 1982), Portugal (Correia 1982), Hungary (Czeizel 1994), Tanzania (Fawzi 1998; Fawzi 2007), Nigeria (Fleming 1968; Fleming 1986), Burkino Faso (Roberfroid 2008), Japan (Hemmi 2003), India (ICMR 2000), Nepal (Christian 2003; Katz 2000; Osrin 2005), the Republic of Ireland (Kirke 1992), Australia (Rumbold 2006), Brazil (Spinnato 2007), Malawi (Kumwenda 2002; Van den Broek 2006), Indonesia (Rumiris 2006; Schmidt 2001; The Summit 2008) and South Africa (Steyn 2003). One trial involved 33 international centres (MRC 1991) and another trial involved India, Peru, South Africa and Viet Nam (Villar 2009).

Interventions

The 28 trials assessed a range of vitamin supplements, alone or in combination with other supplements. The vitamins included vitamin A, alone or with iron, folic acid, zinc or multivitamins (Christian 2003; Fawzi 1998; Katz 2000; Kumwenda 2002; Schmidt 2001; Van den Broek 2006), vitamin C with or without multivitamins or vitamin E (Briscoe 1959; Chappell 1999; Hemmi 2003; Rumbold 2006; Spinnato 2007; Steyn 2003; Villar 2009), folic acid with or without multivitamins and/or iron (Correia 1982; Czeizel 1994; Fleming 1968; Fleming 1986; ICMR 2000; Kirke 1992; MRC 1991; Taylor 1982), multivitamins with iron and folic acid (Fawzi 2007; Osrin 2005; Roberfroid 2008; Rumiris 2006; The Summit 2008) and multivitamins alone (People's League 1942; Rush 1980). The doses of vitamins were similar for the vitamin C supplementation trials (range 400 mg to 1000 mg). However, they varied widely between trials for the folic acid (range 0.3 mg to 10 mg), multivitamins and vitamin A trials (range 5000 international units (IU) to 23,300 IU).

We were unable to include data from 12 trials in the analysis of 'any vitamins versus no or minimal vitamins' because the trials either had more than two treatment arms or compared one or more vitamin interventions with each other. For example, one trial (Fawzi 1998) using a 2 x 2 factorial design compared vitamin A supplements with or without multivitamins versus multivitamins (excluding vitamin A) or placebo. However, results were not explicitly presented for each group. Similarly, three trials (Kumwenda 2002; Schmidt 2001; Van den Broek 2006) compared vitamin A supplements with iron and folic acid versus iron and folic acid alone, and one trial with five treatment arms (Christian 2003) compared multivitamins with iron, folic acid and vitamin A versus iron, folic acid and vitamin A alone. Other trials compared folic acid alone with multivitamins plus folic acid or multivitamins excluding folic acid (Kirke 1992), multivitamins with vitamin E compared with multivitamins without vitamin E (Rush 1980), or multivitamins with iron and folic acid versus iron and folic acid (Fawzi 2007; Osrin 2005; Roberfroid 2008; Rumiris 2006; The Summit 2008). We used data from these trials only in the sub analyses according to vitamin type, as none of the treatment arms were comparable to the control groups used in the trials included in the 'any vitamins versus no or minimal vitamins' comparisons.

Outcomes
Main outcomes

Twenty-four trials reported either pregnancy loss as miscarriage or stillbirth. We included four trials (Christian 2003; Correia 1982; Taylor 1982; Villar 2009) as they reported main outcomes (perinatal death, neonatal death, infant death, preterm birth, birthweight, small-for-gestational age); however, information on miscarriage or stillbirth was either not reported separately or at all. The outcome 'total fetal loss' included both miscarriage or stillbirth, and overcame problems with different definitions of miscarriage and stillbirth. For some trials, miscarriage was considered to occur up until 26 or 28 weeks' gestation, while other studies reported miscarriage as pregnancy loss prior to 20 weeks' gestation, and stillbirth as pregnancy loss greater than or equal to 20 weeks' gestation. Other studies did not specify their definition of miscarriage or stillbirth. For the other main outcomes for the mother, five trials reported placental abruption (Chappell 1999; Rumbold 2006; Rumiris 2006; Spinnato 2007; Villar 2009), and another trial (Steyn 2003) reported antepartum haemorrhage including placental abruption. Eight trials reported pre-eclampsia or 'toxaemia' (Chappell 1999; Fleming 1968; People's League 1942; Rumbold 2006; Rumiris 2006; Spinnato 2007; Steyn 2003; Villar 2009); no trials reported any psychological effects. For the infant, perinatal death was reported in nine trials (Christian 2003; Fawzi 2007; Osrin 2005; Roberfroid 2008; Rumbold 2006; Spinnato 2007; Steyn 2003; The Summit 2008; Villar 2009), neonatal death in 12 trials (Christian 2003; Czeizel 1994; Katz 2000; Fawzi 2007; Osrin 2005; People's League 1942; Roberfroid 2008; Rumbold 2006; Rush 1980; Spinnato 2007; Steyn 2003; The Summit 2008), preterm birth in sixteen trials (Chappell 1999; Christian 2003; Czeizel 1994; Fawzi 1998; Fawzi 2007; Fleming 1968; Katz 2000; Osrin 2005; Roberfroid 2008; Rumiris 2006; Rumbold 2006; Rush 1980; Spinnato 2007; Steyn 2003; The Summit 2008; Van den Broek 2006), very preterm birth in six trials (Fawzi 1998; Fawzi 2007; Rumbold 2006; Spinnato 2007; Steyn 2003; Villar 2009), birthweight in 10 trials (Christian 2003; Correia 1982; Czeizel 1994; Fawzi 2007; Kumwenda 2002; Osrin 2005; Roberfroid 2008; Rumbold 2006; Spinnato 2007; Taylor 1982), small-for-gestational age in 13 trials (Chappell 1999; Christian 2003; Czeizel 1994b; Fawzi 1998; Fawzi 2007; Fleming 1968; ICMR 2000; Roberfroid 2008; Rumbold 2006; Rumiris 2006; Spinnato 2007; The Summit 2008; Villar 2009) and congenital malformations in six trials (Czeizel 1994; Kirke 1992; MRC 1991; Osrin 2005; Spinnato 2007; Villar 2009). In this review, congenital malformations covered malformations excluding neural tube defects, as these are covered in the Cochrane review 'Periconceptional supplementation with folate and/or multivitamins for preventing neural tube defects' (Lumley 2001b).

Other outcomes

Five trials reported multiple pregnancy (Czeizel 1994b; Fleming 1968; ICMR 2000; Katz 2000; Kumwenda 2002); one trial reported Apgar score less than seven at five minutes (Spinnato 2007); six trials reported maternal anaemia (variously defined) (Christian 2003; Fawzi 1998; Fawzi 2007; Fleming 1986; Osrin 2005; Van den Broek 2006); two trials reported infant anaemia at various ages (Fawzi 1998; Kumwenda 2002) and one trial reported placental weight (Correia 1982). Three trials reported various measures of childhood growth including weight and length at six weeks and four months (Kumwenda 2002; Schmidt 2001), and stunting and underweight in children aged 6-8 years (Christian 2003). One trial reported on mode of feeding as the number of women breastfeeding (People's League 1942). One trial reported on adverse effects of vitamin supplementation sufficient to stop supplementation (Spinnato 2007) and one trial reported on side effects (Rumbold 2006). Three trials reported on measures of use of health service resources including maternal admission to ICU (Villar 2009), and admission to the neonatal intensive care unit (Rumbold 2006; Steyn 2003; Villar 2009). The following outcomes were not reported by any of the trials: use of blood transfusion for the mother, subsequent fertility, disability at childhood follow up, and maternal views of care.

Additional outcomes

One trial (Steyn 2003) reported birthweight as the median birthweight and range; however, these data were not in a format suitable for inclusion in the birthweight comparisons (see Table 1). One trial (Christian 2003) reported infant death and we have included this outcome in the 'other outcomes reported' for the multivitamin and folic acid comparisons.

Table 1. Additional outcomes
Study IDOutcomeVitamin - NVitamin - MedianVitamin - RangePlacebo - NPlacebo - MedianPlacebo - Range
Steyn 2003Median birthweight1002491240-38341002664334-4680

Excluded studies

We excluded 38 trials, of which 12 reported no clinically meaningful data in a format suitable for inclusion (Hibbard 1968; Laurence 1981; Lira 1989; Meirinho 1987; Mock 2002; Moldenhauer 2002; Semba 2001; Suharno 1993; Tanumihardjo 2002; Thauvin 1992; Villamor 2002; Vutyavanich 1995). Six trials did not clearly report the gestational age when supplementation was started (Biswas 1984; Fletcher 1971; Hampel 1974; Lumeng 1976; Schuster 1984; Trigg 1976) and for two trials the majority of women were enrolled after 20 weeks and did not report outcomes separately for women starting supplementation prior to 20 weeks (Ferguson 1955; Giles 1971). Eleven trials (Baumslag 1970; Blot 1981; Chanarin 1968; Colman 1974; Coutsoudis 1999; Dawson 1962; Edelstein 1968; Feyi-Waboso 2005; Hankin 1966; Marya 1981; Metz 1965; Owen 1966) reported supplementation after 20 weeks' gestation. One trial (Ross 1985) did not specify the contents of the supplements; in three trials all women were given a vitamin supplement (Hunt 1984; Huybregts 2009; Shu 2002); and two were non-randomised (Smithells 1981; Ulrich 1999).

Three other trials (Beazley 2002; Chaudhuri 1969; Rivas 2000) supplemented women for the prevention of pre-eclampsia, and did not report any outcomes related to pregnancy loss. These trials are covered in the Cochrane review 'Antioxidants for preventing pre-eclampsia' (Rumbold 2008).

Risk of bias in included studies

Figure 1 and Figure 2 illustrate that the trials were of variable quality. In the main analysis, only 10% to 20% weight is from high-quality trials. Two trials (Fleming 1968; People's League 1942) used quasi-random allocation methods involving alternate allocation of participants. Similarly, three trials (Christian 2003; Katz 2000; The Summit 2008) used cluster randomisation. Two of these (Christian 2003; Katz 2000) allocated subdistricts within Nepal to each treatment arm and randomised women by "drawing numbered identical chits from a hat, blocked on subdistrict"; one trial (The Summit 2008) allocated individual midwives to each treatment arm. Concealment of allocation was adequate in 13 trials and blinding was adequate in 17 trials. For many of the other trials there was inadequate reporting of methodological details to make an assessment about the risk of bias.

Figure 1.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies

Figure 2.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study

In order to examine possible publication bias, we undertook funnel plots which graph the effect size against sample size for each trial for outcomes with 10 or more studies. For the comparisons of any vitamins versus no vitamins, the funnel plots were symmetrical for the outcomes total fetal loss (Figure 3) and early or late miscarriage (Figure 4), indicating a low possibility of publication bias. The graphs for the analysis of multivitamins were less symmetrical for the outcomes total fetal loss (Figure 5), early or late miscarriage (Figure 6), indicating the possibility that smaller negative trials may be missing.

Figure 3.

Funnel plot of comparison: 1 Any vitamins versus no vitamins (or minimal vitamins), outcome: 1.1 Total fetal loss (including miscarriages or combined miscarriages and stillbirths)

Figure 4.

Funnel plot of comparison: 1 Any vitamins versus no vitamins (or minimal vitamins), outcome: 1.2 Early or late miscarriage

Figure 5.

Funnel plot of comparison: 5 Multivitamin, outcome: 5.1 Total fetal loss (including miscarriages or combined miscarriages and stillbirths)

Figure 6.

Funnel plot of comparison: 5 Multivitamin, outcome: 5.2 Early or late miscarriage

Effects of interventions

See: Summary of findings for the main comparison Vitamin A versus placebo for preventing miscarriage; Summary of findings 2 Vitamin A versus B-Carotene for preventing miscarriage; Summary of findings 3 Vitamin A plus iron plus folate versus iron plus folate for preventing miscarriage

We have included 28, involving 96,674 women and 98,267 pregnancies.

Any vitamins versus no vitamins (or minimal vitamins)

Primary outcomes
For the woman

For the outcomes of miscarriage and stillbirth, 13 trials contributed data which included 33,943 pregnancies (Briscoe 1959; Chappell 1999; Czeizel 1994; Fleming 1968; Fleming 1986; Hemmi 2003; ICMR 2000; Katz 2000; MRC 1991; People's League 1942; Rumbold 2006; Spinnato 2007; Steyn 2003). No difference was seen between women given any type of vitamin(s) compared with controls for total fetal loss (relative risk 1.04, 95% confidence interval (CI) 0.95 to 1.14, Analysis 1.1) or early or late miscarriage (risk ratio (RR) 1.09, 95% CI 0.95 to 1.25, Analysis 1.2), using fixed-effect models. These findings occurred regardless of whether the trials started supplementation prior to pregnancy, in the first 12 weeks of pregnancy, before 20 weeks' gestation or both prior to and after 20 weeks' gestation. For the other primary maternal outcomes, no clear difference was seen between women given any type of vitamin(s) compared with controls for placental abruption (RR 0.66, 95% CI 0.34 to 1.30, four trials, 4264 women (Chappell 1999; Rumbold 2006; Spinnato 2007; Villar 2009), Analysis 1.3), antepartum haemorrhage including placental abruption (RR 7.00, 95% CI 0.88 to 55.86, one trial, 200 women (Steyn 2003)).

Heterogeneity

We found significant heterogeneity for pre-eclampsia. For pre-eclampsia, this heterogeneity is likely to be due to the inclusion of People's League 1942, which was quasi-randomised. Using a random-effects model, there was no clear difference between women given any type of vitamin(s) compared with controls for pre-eclampsia (RR 0.88, 95% CI 0.70 to 1.09, 7 trials, 9561 women, Tau2 0.04 (Chappell 1999; Fleming 1968; People's League 1942; Rumbold 2006; Spinnato 2007; Steyn 2003; Villar 2009), Analysis 1.5).

For the infant

No difference was seen between women given any type of vitamin(s) compared with controls for stillbirth (RR 0.86, 95% CI 0.65 to 1.13, nine trials, 15,980 women (Chappell 1999; Czeizel 1994; Fleming 1968; ICMR 2000; MRC 1991; People's League 1942; Rumbold 2006; Spinnato 2007; Steyn 2003), Analysis 1.6); perinatal death (RR 0.83, 95% CI 0.62 to 1.11, four trials, 4313 women (Rumbold 2006; Spinnato 2007; Steyn 2003; Villar 2009), Analysis 1.7); neonatal death (relative risk 1.11, 95% CI 0.94 to 1.31, six trials, 27,657 women (Czeizel 1994b; Katz 2000; People's League 1942; Rumbold 2006; Spinnato 2007; Steyn 2003), Analysis 1.8); preterm birth (relative risk 1.02 95% CI 0.94 to 1.10, eight trials, 27,657 women (Chappell 1999; Czeizel 1994; Fleming 1968; Katz 2000; Rumbold 2006; Spinnato 2007; Steyn 2003; Villar 2009), Analysis 1.9); small-for-gestational-age infants (RR 0.96 95% CI 0.84 to 1.08, seven trials, 9,356 women (Chappell 1999; Czeizel 1994; Fleming 1968; ICMR 2000; Rumbold 2006; Spinnato 2007; Villar 2009), Analysis 1.12) or congenital malformations (RR 1.47, 95% CI 0.90 to 2.40, four trials, 8933 women (Czeizel 1994; MRC 1991; Spinnato 2007; Villar 2009), Analysis 1.13).

Heterogeneity

We found significant heterogeneity for birthweight. This is likely to be due to the inclusion of Correia 1982, which was at high risk of bias due to more than 20% of participants being excluded. Using a random-effects model, there was no clear difference between women given any type of vitamin(s) compared with controls for birthweight (mean difference (MD) 16.99 g, 95% CI -37.66 to 71.64, five trials, 7497 women, Tau2 1748.69 (Correia 1982; Czeizel 1994b; Rumbold 2006; Spinnato 2007; Taylor 1982), Analysis 1.11). When this trial was excluded from the analysis there was no heterogeneity present; however the direction of effects did not change (MD 2.01, 95% CI -21.51-25.52, four trials, 7468 women (Czeizel 1994b; Rumbold 2006; Spinnato 2007; Taylor 1982)).

Secondary outcomes

Women given any type of vitamin(s) compared with controls were more likely to have a multiple pregnancy (relative risk 1.38, 95% CI 1.12 to 1.70, three trials, 20,986 women (Czeizel 1994b; ICMR 2000; Katz 2000)) and greater placental weight (MD 96.00 g, 95% CI 30.73 to 161.27, one trial, 29 women (Correia 1982)). No overall difference was seen between women given any vitamin(s) compared with controls for the outcomes very preterm birth (RR 0.93, 95% CI 0.75 to 1.15, four trials, 4181 women (Rumbold 2006; Spinnato 2007; Steyn 2003; Villar 2009)); Apgar score less than seven at five minutes (RR 0.66, 95% CI 0.27 to 1.60, 1 trial, 700 women (Spinnato 2007)); infant anaemia (RR 1.05, 95% CI 0.98 to 1.12, one trial, 836 infants (Fawzi 1998)); the number of women breastfeeding (RR 0.98, 95% CI 0.96 to 1.01, one trial, 4878 women (People's League 1942)); any maternal admission to the intensive care unit (RR 0.20, 95% CI 1.69, one trial, 1365 women (Villar 2009)); any admission to the neonatal intensive care unit (RR 0.81, 95% CI 0.59 to 1.11, one trial, 1515 infants (Villar 2009)) or the duration of admission to the neonatal intensive care unit (MD 1.30 days, 95% CI -0.28 to 2.88, one trial, 181 women (Steyn 2003)). There was no significant difference between women given any type of vitamin(s) compared with controls for adverse effects sufficient to stop supplementation (RR 1.16, 95% CI 0.39 to 3.41, one trial, 739 women (Spinnato 2007)); however, women given any type of vitamin(s) were more likely to report abdominal pain compared with controls (RR 1.63, 95% CI 1.12 to 2.36, one trial, 1734 women (Rumbold 2006)).

Heterogeneity

We found significant heterogeneity for maternal anaemia. The cause of this heterogeneity is unclear, however it may be due to differences in the participants; for example, one study (Fawzi 1998) included women who were HIV positive. There was no significant difference between women given any type of vitamin(s) compared with controls for maternal anaemia (RR 0.90, 95% CI 0.46 to 1.73, 2 trials, 1190 women, Tau2 0.16 (Fawzi 1998; Fleming 1986)) using a random-effects model. No other secondary outcomes were reported.

Sensitivity analyses by quality rating

Five trials had an allocation concealment rating of 'Adequate' (Chappell 1999; Kirke 1992; Kumwenda 2002; MRC 1991; Steyn 2003); 10 trials had a rating of 'Unclear' (Briscoe 1959; Christian 2003; Correia 1982; Czeizel 1994; Fawzi 1998; Hemmi 2003; ICMR 2000; Katz 2000; Rush 1980; Schmidt 2001) and two trials had a rating of 'No' (Fleming 1968; People's League 1942). The sensitivity analyses excluded the trials with an allocation concealment rating of unclear or 'No' (inadequate). Amongst the trials with adequate allocation concealment (high quality studies), there was no difference in total fetal loss between women supplemented with any vitamins compared with controls (relative risk 0.97, 95% CI 0.84 to 1.12, five trials, 4916 women (Chappell 1999; MRC 1991; Rumbold 2006; Spinnato 2007; Steyn 2003)). Similarly, we found no difference between women supplemented with any vitamins compared with controls for early or late miscarriage or stillbirth, when the analyses were restricted to high quality studies only. These sensitivity analyses indicate that the analyses for the effects of any vitamins on outcomes related to fetal loss are no different when only high quality studies are included.

Subgroup analyses by vitamin type

Vitamin C supplementation

The trials involving vitamin C supplementation included the following interventions: vitamin C plus multivitamins versus placebo plus multivitamins (Briscoe 1959), vitamin C and vitamin E supplementation versus placebo (Chappell 1999; Rumbold 2006; Spinnato 2007; Villar 2009) and vitamin C alone versus no supplement or placebo (Hemmi 2003; Steyn 2003).

Primary outcomes
For the woman

We found no significant difference in the risk of total fetal loss between women receiving vitamin C with multivitamins compared with placebo plus multivitamins (RR 1.32, 95% CI 0.63 to 2.77, one trial, 406 women (Briscoe 1959)), vitamin C and vitamin E compared with placebo (RR 0.82, 95% CI 0.48 to 1.42, three trials, 2899 women (Chappell 1999; Rumbold 2006; Spinnato 2007)) or vitamin C compared with no supplement or placebo (RR 1.28, 95% CI 0.58 to 2.83, two trials, 224 women (Hemmi 2003; Steyn 2003)) (Analysis 3.1). Similarly, there was no overall difference seen in early or late miscarriage between women receiving vitamin C with multivitamins compared with placebo plus multivitamins (RR 1.32, 95% CI 0.63 to 2.77, one trial, 406 women (Briscoe 1959)), vitamin C and vitamin E compared with placebo (RR 0.70, 95% CI 0.27 to 1.84, two trials, 2616 women (Rumbold 2006; Spinnato 2007) or vitamin C compared with no supplement or placebo (RR 1.17, 95% CI 0.52 to 2.65, two trials, 224 women (Hemmi 2003; Steyn 2003)).

There was no significant difference between women receiving any combination of vitamin C supplementation compared with control for other primary outcomes including placental abruption or antepartum haemorrhage.

Heterogeneity

We found significant heterogeneity for the outcome pre-eclampsia in the subgroup analyses of vitamin C and E versus placebo. Using a random-effects analysis, there was no significant difference in the risk of pre-eclampsia amongst women given vitamin C and E supplementation compared with a placebo (RR 0.94, 95% CI 0.72 to 1.22, four trials, 4264 women, Tau2 0.04 (Chappell 1999; Rumbold 2006; Spinnato 2007; Villar 2009)), or women given vitamin C alone compared with placebo (RR 1.00, 95% CI 0.21 to 4.84, one trial, 200 women (Steyn 2003)). The heterogeneity appears to be due to the inclusion of Chappell 1999, and may be due to differences in the risk profile of participants. When we excluded this study from the analysis there was no heterogeneity, and the direction of effects moved towards the null (RR 1.03, 95% CI 0.88 to 1.20, three trials, 4081 women (Rumbold 2006; Spinnato 2007; Villar 2009)).

For the infant

We found no significant difference in the risk of stillbirth between women receiving vitamin C and vitamin E compared with placebo (RR 0.89, 95% CI 0.46 to 1.73, three trials, 2899 women (Chappell 1999; Rumbold 2006; Spinnato 2007)) or vitamin C compared with placebo (RR 3.00, 95% CI 0.12 to 72.77, one trial, 200 women (Steyn 2003)). There was no significant difference between women receiving any combination of vitamin C supplementation compared with control for other primary outcomes including perinatal death, neonatal death, very preterm birth (less than 34 weeks), small-for-gestational-age infants, birthweight and congenital malformations.

Women supplemented with vitamin C compared with placebo were at increased risk of having a preterm birth (RR 1.43, 95% CI 1.03 to 1.99, one trial, 200 women (Steyn 2003)); however, no increased risk of preterm birth was seen when women were given vitamin C in addition to vitamin E compared with placebo (RR 0.97, 95% CI 0.85 to 1.10, four trials, 4264 women (Chappell 1999; Rumbold 2006; Spinnato 2007; Villar 2009)).

Secondary outcomes

There was no significant difference between women receiving any combination of vitamin C supplementation compared with control for the secondary outcomes: Apgar score less than seven at five minutes, maternal hospitalisation or admission to the neonatal intensive care unit. There was no significant difference between women given vitamin C and vitamin E compared with placebo for adverse effects sufficient to stop supplementation (RR 1.16, 95% CI 0.39 to 3.41, one trial, 739 women (Spinnato 2007)); however, women given vitamin C and vitamin E were more likely to report abdominal pain compared with women given a placebo (RR 1.63, 95% CI 1.12 to 2.36, one trial, 1734 women (Rumbold 2006)).

Vitamin A supplementation

The trials involving vitamin A supplementation included the following interventions: vitamin A and/or beta-carotene versus placebo (Katz 2000), vitamin A with or without multivitamins versus multivitamins (excluding vitamin A) or placebo (Fawzi 1998) and vitamin A plus iron and folic acid versus iron and folic acid (Kumwenda 2002; Schmidt 2001; Van den Broek 2006).

Primary outcomes
For the woman

We found no difference in total fetal loss between women given vitamin A compared with placebo (relative risk 1.04, 95% CI 0.92 to 1.17, one trial, 11,723 women (Katz 2000)), beta-carotene compared with placebo (RR 1.03, 95% CI 0.91 to 1.16, one trial, 11,303 women (Katz 2000)), vitamin A compared with beta-carotene (RR 1.01, 95% CI 0.90 to 1.14, one trial, 11,720 women (Katz 2000)), vitamin A or beta-carotene compared with placebo (RR 1.05, 95% CI 0.91 to 1.21, one trial, 17,373 women (Katz 2000)), vitamin A with or without multivitamins compared with multivitamins or placebo (RR 0.80, 95% CI 0.53 to 1.21, one trial, 1074 women (Fawzi 1998)) or vitamin A with iron and folic acid compared with iron and folic acid (RR 1.01, 95% CI 0.61 to 1.66, three trials, 1640 women (Kumwenda 2002; Schmidt 2001; Van den Broek 2006)) (Analysis 4.1). Similarly, we found no differences in the rate of early or late miscarriage.

For the infant

We found no differences in the rate of stillbirth, neonatal death, preterm birth or very preterm birth between women given any type of vitamin A, alone or in combination with beta-carotene, multivitamin or iron and folic acid, compared with controls. Infants of women given vitamin A with iron and folic acid compared with iron and folic acid alone had higher birthweight (MD 90.00 g, 95% CI 2.68 to 177.32, one trial, 594 women (Kumwenda 2002)). However, we found no difference in the number of infants born small-for-gestational age between women given vitamin A with or without multivitamins compared with multivitamins or placebo (RR 0.84, 95% CI 0.58 to 1.21, one trial, 1075 women (Fawzi 1998)).

Secondary outcomes

The rate of multiple pregnancy was higher in women given either vitamin A or beta-carotene compared with placebo (relative risk 1.39, 95% CI 1.05 to 1.84, one trial, 15,845 women (Katz 2000)), and there were trends towards an increase in the rate of multiple pregnancy for women given vitamin A versus placebo (relative risk 1.35, 95% CI 0.99 to 1.85, one trial, 10,697 women (Katz 2000)) or beta-carotene versus placebo (relative risk 1.37, 95% CI 1.00 to 1.88, one trial, 10,294 women (Katz 2000)). Fewer infants of women given vitamin A and iron and folic acid compared with iron and folic acid alone had anaemia at six weeks of age (RR 0.58, 95% CI 0.45 to 0.75, one trial, 562 infants (Kumwenda 2002)); however, at 12 months, no difference was seen (RR 1.03, 95% CI 0.88 to 1.20, one trial, 478 infants (Kumwenda 2002)). Similarly there was no difference in infant anaemia amongst women given other combinations of vitamin A during pregnancy, and no differences in the rate of maternal anaemia for women given any combination of vitamin A in pregnancy. At six weeks of age, infants of women given vitamin A and iron and folic acid compared with iron and folic acid alone had greater weight (MD 169 g, 95% CI 16.55 to 321.45, one trial, 546 infants (Kumwenda 2002)) and length (MD 0.70 cm, 95% CI 0.15 to 1.25, one trial, 546 infants (Kumwenda 2002)); however, at four months of age we found no significant difference in weight (MD -100.00 g, 95% CI -377.14 to 177.14, one trial, 148 infants (Schmidt 2001)) or length (MD -0.50 cm, 95% CI -1.33 to 0.33, one trial, 148 infants (Schmidt 2001)).

Multivitamin supplementation

The trials involving multivitamin supplementation included the following interventions: multivitamins with or without folic acid versus no multivitamins or folic acid (Czeizel 1994; MRC 1991); multivitamins with or without folic acid versus folic acid (Kirke 1992; MRC 1991); multivitamins with or without vitamin A versus vitamin A or placebo (Fawzi 1998); multivitamins versus control (People's League 1942); multivitamins with vitamin E versus multivitamins without vitamin E or control (Rush 1980); multivitamins with iron and folic acid versus iron and folic acid (Fawzi 2007; Osrin 2005; Roberfroid 2008; Rumiris 2006; The Summit 2008) and multivitamins with folic acid, iron, zinc and vitamin A versus no multivitamin and folic acid, iron, zinc, vitamin A (Christian 2003).

Primary outcomes
For the woman

We found moderate heterogeneity in the subgroup analyses of total fetal loss, early or late miscarriage and stillbirth amongst women given multivitamins in various combinations with iron and folic acid. This heterogeneity may be explained by differences in study populations, for example the inclusion of Rumiris 2006, which included women with low antioxidant status at entry and ICMR 2000, which included women who had previously given birth to a child with an open neural tube defect. Total fetal loss was significantly lower in women who were given multivitamins with or without vitamin A compared with those receiving vitamin A or placebo (RR 0.60, 95% CI 0.39 to 0.91, one trial, 1074 women (Fawzi 1998)), using a random-effects analysis (Analysis 5.1). However, given that these findings occurred in subgroup analyses and they include studies where the comparison groups contain either women receiving either vitamin A or placebo, caution should be taken in interpretation. Using random-effects analyses, we found no other differences in total fetal loss for women receiving multivitamins with folic acid versus no multivitamins or folic acid (RR 1.00, 95% CI 0.75 to 1.34, three trials, 6883 women, Tau2 0.03 (Czeizel 1994; ICMR 2000; MRC 1991)); multivitamins without folic acid versus no multivitamins or folic acid (RR 0.83, 95% CI 0.56 to 1.25, one trial, 907 women (MRC 1991)); multivitamins with or without folic acid versus no multivitamins or folic acid (RR 0.91, 95% CI 0.65 to 1.27, one trial, 1368 women (MRC 1991)); multivitamins with folic acid versus folic acid (RR 1.03, 95% CI 0.72 to 1.48, two trials, 1096 women (Kirke 1992; MRC 1991)); multivitamins without folic acid versus folic acid (RR 0.90, 95% CI 0.62 to 1.30, two trials, 1090 women (Kirke 1992; MRC 1991)); multivitamins with or without folic acid versus folic acid (RR 0.95, 95% CI 0.69 to 1.30, two trials, 1644 women (Kirke 1992; MRC 1991)); multivitamins versus control (RR 0.83, 95% CI 0.58 to 1.17, one trial, 5021 women (People's League 1942)); multivitamins with vitamin E versus multivitamins without vitamin E or control (RR 0.92, 95% CI 0.46 to 1.83, one trial, 823 women (Rush 1980)) and women given multivitamins with iron and folic acid versus iron and folic acid (RR 0.90, 95% CI 0.75 to 1.09, five trials, 42,404 women, Tau2 0.02 (Fawzi 2007; Osrin 2005; Roberfroid 2008; Rumiris 2006; The Summit 2008)). Using random-effects models, there was no clear difference in the risk of early or late miscarriage between women supplemented with multivitamins compared with control, for any of the multivitamin comparisons (Analysis 5.2).

For the other primary outcomes, women receiving multivitamins compared with control were at lower risk of pre-eclampsia (RR 0.70, 95% CI 0.55 to 0.90, one trial, 5021 women (People's League 1942)), and there was a trend to a reduced risk of pre-eclampsia amongst women receiving multivitamins with iron and folic acid compared with iron and folic acid (RR 0.24, 95% CI 0.06 to 1.01, one trial, 60 women, (Rumiris 2006)), which was of borderline statistical significance (P = 0.05).

For the infant

There was no clear difference in the risk of stillbirth for any of the multivitamin comparisons using random-effects models (Analysis 5.5). Infants of women given multivitamins with iron and folic acid compared with iron and folic acid had a higher birthweight (MD 61.61 grams, 95% CI 37.32 to 85.91, three trials, 10241 infants (Fawzi 2007; Osrin 2005; Roberfroid 2008)). There were no clear differences demonstrated in the risk of neonatal death, infant death, preterm birth, congenital malformations.

Heterogeneity

In addition to total fetal loss, early or late miscarriage and stillbirth (all discussed above), we also identified substantial heterogeneity for perinatal death and small-for-gestational age in the comparisons of women given multivitamins with iron and folic acid compared with iron and folic acid alone. The source of this heterogeneity is unclear; it may be due to different characteristics of women in these studies. Using a random-effects model, there were no clear differences in the risks of perinatal death or small-for-gestational age between any of the multivitamin combinations compared with controls, with the exception of women given multivitamins with folic acid, iron, zinc and vitamin A. In this group there was a trend towards a reduction in the number of infants with a birthweight less than 2500 g amongst women given multivitamins with folic acid, iron, zinc and vitamin A versus no multivitamin and folic acid, iron, zinc, vitamin A (relative risk 0.95 95% CI 0.90 to 1.00, one trial, 3325 infants (Christian 2003)).

Secondary outcomes

There was no clear difference in the risk of maternal anaemia or in the number of women breastfeeding, for any of the multivitamin comparisons. Women given multivitamins with folic acid compared with no multivitamins or folic acid were more likely to have a multiple pregnancy (RR 1.36, 95% CI 1.00 to 1.85, two trials, 5141 women (Czeizel 1994b; ICMR 2000)), although this result was of borderline statistical significance (P = 0.05). Women given multivitamins with iron and folic acid had a reduced risk of anaemia compared with women given iron and folic acid alone (RR 0.88, 95% CI 0.81 to 0.96, two trials, 2278 women (Fawzi 2007; Osrin 2005)). During childhood follow-up, children of women given multivitamins with folic acid, iron, zinc and vitamin A versus no multivitamin and folic acid, iron, zinc, vitamin A were more likely to be stunted at six to eight years of age (relative risk 1.09, 95% CI 1.00 to 1.19, one trial, 3356 children (Christian 2003)), although this result was of borderline statistical significance (P = 0.05).

Folic acid supplementation

The trials involving folic acid supplementation included the following interventions: folic acid with or without multivitamins compared with no folic acid or multivitamins (Czeizel 1994; ICMR 2000; MRC 1991); folic acid with or without multivitamins compared with multivitamins (Kirke 1992; MRC 1991); folic acid with iron, zinc, multivitamins and vitamin A compared with vitamin A alone (Christian 2003); folic acid and iron compared with iron (Fleming 1968); folic acid and iron compared with no iron or folic acid (Fleming 1986) and folic acid compared with placebo (Correia 1982).

Primary outcomes
For the woman

We found no significant difference in the risk of total fetal loss between women receiving folic acid with multivitamins compared with no folic acid or multivitamins (RR 1.09, 95% CI 0.95 to 1.25, three trials, 6883 women (Czeizel 1994; ICMR 2000; MRC 1991)); folic acid without multivitamins compared with no folic acid or multivitamins (RR 0.95, 95% CI 0.64 to 1.40, one trial, 903 women (MRC 1991)); folic acid with or without multivitamins compared with no folic acid or multivitamins (RR 0.97, 95% CI 0.69 to 1.35, one trial, 1364 women (MRC 1991)); folic acid with multivitamins compared with multivitamins (RR 1.15, 95% CI 0.80 to 1.67, two trials, 1102 women (Kirke 1992; MRC 1991)); folic acid without multivitamins compared with multivitamins (RR 1.12, 95% CI 0.77 to 1.62, two trials, 1090 women (Kirke 1992; MRC 1991)); folic acid with or without multivitamins compared with multivitamins (RR 1.14, 95% CI 0.82 to 1.57, two trials, 1644 women (Kirke 1992; MRC 1991)); folic acid with iron compared with iron (RR 0.23, 95% CI 0.01 to 4.59, one trial, 75 women (Fleming 1968)) and folic acid and iron compared no iron or folic acid (RR 13.00, 95% CI 0.74 to 226.98, one trial, 160 women (Fleming 1986)) (Analysis 6.1).

For the other primary outcomes, we found no differences in the risk of pre-eclampsia between women given folic acid, alone or with multivitamins or iron, compared with controls.

Heterogeneity

We identified significant heterogeneity in the subgroup analyses of early or late miscarriage comparing folic acid and multivitamins versus no folic acid or multivitamins (Tau2 0.04). This is likely to be due to the inclusion of ICMR 2000, which included women who had previously given birth to a child with an open neural tube defect and was at increased risk of bias due to incomplete outcome data for all participants. There was no clear difference in the risk of early or late miscarriage between women given folic acid compared with control, for any of the folic acid combinations (Analysis 6.2). Excluding the ICMR study from the analysis reduced the heterogeneity, and did not change the overall direction of effects.

For the infant

We found no differences in the risk of stillbirth, perinatal death, neonatal death, preterm birth, small-for-gestational-age infants (birthweight less than 10th centile), congenital malformations or infant death between women given folic acid, alone or with multivitamins or iron, compared with controls. Women given folic acid compared with placebo had a greater placental weight (MD 96 g, 95% CI 30.73 to 161.27, one trial, 29 women (Correia 1982)), their infants had greater birthweight (MD 312 g, 95% CI 108.52 to 515.48, one trial, 29 women (Correia 1982)); however, we found no difference in birthweight between women given folic acid and multivitamins compared with no folic acid or multivitamins (MD 3.00 g, 95% CI -24.15 to 30.15, one trial, 4862 women (Czeizel 1994)). Women receiving folic acid with iron, zinc, multivitamins and vitamin A compared with vitamin A alone were less likely to have an infant with a birthweight less than 2500 g (relative risk 0.94, 95% CI 0.90 to 0.99, one trial, 3325 women (Christian 2003)); however, we found no differences in birthweight less than 2500 g for women receiving any of the other folic acid comparisons.

Secondary outcomes

Women receiving folic acid and multivitamins compared with no folic acid or multivitamins were more likely to have a multiple pregnancy (RR 1.36, 95% CI 1.00 to 1.85, two trials, 5141 women (Czeizel 1994b; ICMR 2000)). Fewer women given folic acid with iron, zinc, multivitamins and vitamin A compared with vitamin A alone had anaemia in the third trimester (relative risk 0.83, 95% CI 0.77 to 0.91, one trial, 813 women (Christian 2003)); however, there were no differences seen for any of the other folic acid comparisons for maternal anaemia overall and severe maternal anaemia. During childhood follow-up, children of women given folic acid, iron, zinc, vitamin A compared with multivitamins and vitamin A were less likely to be stunted at six to eight years of age (relative risk 0.93, 95% CI 0.86, 1.00, one trial, 3356 children (Christian 2003)).

Subgroup analyses by women's risk of spontaneous or recurrent miscarriage

Information enabling women to be classified at high or low risk of either spontaneous miscarriage or recurrent miscarriage was not clearly stated in any of the trials. Based on the inclusion criteria, one trial (Rumbold 2006) included women at low risk of miscarriage. One trial (Briscoe 1959) included women who had experienced recurrent miscarriage as well as women at high risk of miscarriage (more than two previous miscarriages and/or bleeding in the pregnancy) and low-risk women (two or less previous miscarriages and no bleeding in the pregnancy). After classifying women into these groups, the number of women in the high-risk group was too small to permit any meaningful comparisons and we have therefore not performed subgroup analyses.

Subgroup analyses by dose of vitamins and duration of vitamin usage

Subgroup analyses by dose of vitamin(s) (below or above the recommended dietary intake) were complicated by the limited number of studies in each vitamin group, and by the use of multivitamin supplements. For many of the vitamin types and for those reporting pregnancy loss outcomes, all of the trials supplemented women with amounts that were above the recommended dietary intake. Similarly, the duration of vitamin usage was complicated by the fact that many of the trials had wide recruitment periods, and one trial (Katz 2000) supplemented women up until three years postpartum. Nevertheless, subgroups based on time of trial entry are covered in the main analyses for any vitamins (Analysis 1.1). We have not performed subgroup analyses based on vitamin dosage.

Subgroup analyses by dietary intake of vitamins

Five trials (Fleming 1968; Kumwenda 2002; People's League 1942; Schmidt 2001; Steyn 2003) reported information about women's nutritional status or the percentage of women that were dietary deficient at trial entry for the vitamin of interest. Other trials reported that they were being undertaken in countries where the population was at high risk of multiple micronutrient deficiencies (Osrin 2005; Roberfroid 2008; The Summit 2008; Villar 2009) or anaemia (Fleming 1986), but provided no specific information on nutritional status of participants. Another trial (Rumiris 2006) included women with 'low antioxidant status'. There were not enough trials within each vitamin group to assess the role of supplementation in women with dietary deficient intakes of the individual vitamins and results were not reported separately for women with a low dietary vitamin intake; therefore, we could not perform subgroup analyses.

Discussion

We did not find any evidence to support the use of any vitamins for preventing either early or late miscarriage or stillbirth. Women given multivitamins with or without vitamin A compared with those receiving vitamin A alone or placebo and women given multivitamins with iron and folic acid compared with those receiving iron and folic acid alone were at lower risk of 'total fetal loss'. However, there should be caution interpreting these findings, as they occurred in additional analyses by type of vitamin; the control group involved either vitamin A supplementation, iron and folic acid supplementation or placebo; and findings were no longer significant when total fetal loss was examined separately as miscarriage or stillbirth.

Women given any vitamins alone or in combination with other vitamins during the pre- or peri-conceptional period were more likely to have a multiple pregnancy. These differences persisted in the additional analyses by type of vitamin, whereby vitamin A, multivitamins and folic acid were found to be associated with an increase in multiple pregnancy. These findings are in agreement with the Cochrane review 'Periconceptional supplementation with folic acid and/or multivitamins for preventing neural tube defects' (Lumley 2001b), as well as a re-analysis of individual trial data (Czeizel 1994b) and several other large cohort studies on the rate of twinning amongst users of multivitamins and folic acid (Ericson 2001), and after food fortification with folic acid (Waller 2003). The potential mechanisms of action behind the vitamins and their impact on multiple pregnancy are not well understood, although there is some speculation that the vitamins may influence the rate of twinning rather than improving the survival of multiple fetuses (Katz 2001). An increase in multiple pregnancy along with increases in perinatal morbidity and mortality is of concern; however, as direct causality is yet to be established, further monitoring of pre- and peri-conceptional vitamin supplementation is required. Furthermore, although seven trials in this review enrolled women prior to conception, only three of these trials reported on multiple pregnancy, so there is also potential for bias due to selective reporting of multiple pregnancy.

We detected significant heterogeneity for pre-eclampsia and birthweight in the comparisons of 'any vitamin versus no or minimal vitamins'. This heterogeneity appears to be due to: (a) the inclusion of two studies at high risk of bias due to quasi-random allocation and large losses to follow-up (Correia 1982; People's League 1942); and (b) combining all trials regardless of vitamin type in the analyses. When we conducted subgroup analyses by individual vitamin type, heterogeneity was no longer apparent for birthweight, with the exception of sub-analyses of folic acid supplementation where data were contributed by one trial at high risk of bias (Correia 1982).

In the analyses by vitamin type, supplementing women with vitamin A, iron and folic acid or multivitamins combined with minerals and folic acid was associated with a significant increase in infant birthweight, and for multivitamins, a reduced risk of low birthweight babies, and maternal anaemia. These findings are consistent with the Cochrane review 'Multiple-micronutrient supplementation for women during pregnancy' (Haider 2006), which found a decrease in the number of low birthweight babies, small-for-gestational-age babies and in maternal anaemia amongst women supplemented with multiple micronutrients compared with two or fewer micronutrients or control. However, as the beneficial effects seen for maternal anaemia and infant growth measures were not consistent across all multivitamin combinations, further research is required to confirm the effects of multivitamins on maternal and infant health outcomes. Small improvements in infant growth were also seen with vitamin A and folic acid supplementation. However, these findings should be interpreted with caution, as in many instances the data are contributed from single trials.

Vitamin C supplementation was associated with a small increase in the risk of preterm birth. One trial, involving 200 women at high risk of giving birth preterm, contributed data for this outcome. The authors of this trial concluded that the increase in preterm birth in the vitamin group did not translate into poorer neonatal outcomes (Steyn 2003). Further studies are required before any recommendations can be made regarding vitamin C supplementation. Women supplemented with multivitamins had a reduced risk of pre-eclampsia. However this finding should be interpreted with caution as the data come from one quasi-randomised trial (People's League 1942). The role of vitamins in the prevention of pre-eclampsia is explored in the Cochrane review 'Antioxidants for preventing pre-eclampsia' (Rumbold 2008).

Our review included trials that randomised women prior to conception; however, in some cases, not all women enrolled in these trials fell pregnant during the study period. Some of the trials reported outcomes only for women falling pregnant, whereas other trials did not distinguish between women that were never pregnant and women that may have been pregnant but were lost to follow-up. The outcomes in this review relating to pregnancy complications are not relevant for the women that never became pregnant during the study period. In this review, where trials provided accurate information about the number of women who joined the study and became pregnant in the time period, we included this number in the totals, rather than the number of women that may have been randomised. Where it was not clear about the exact number of women with a confirmed pregnancy, we included all women that had been randomised. This may therefore mean that women in the denominator were never pregnant during the study period. By including these women who were never pregnant in the totals, the review assumes that if these women had become pregnant, they would not have had a miscarriage, which is unlikely to be entirely correct. Including these women creates the potential to underestimate any treatment effects observed.

Similarly, for one large trial (Katz 2000) and one smaller trial (Roberfroid 2008), some women were pregnant more than once during the study period. In these trials, the denominators reported are the total number of pregnancies during the study period, not the total number of women randomised, which incorrectly assumes that each data point included is independent from the next. This has the potential to either underestimate or overestimate the results, depending on whether the women contributing data for more than one pregnancy may be more or less susceptible to experiencing miscarriage or stillbirth. One way to overcome this may be to summarise the data for each woman so that there is only one set of data points for each woman; however, we were unable to do this for these particular studies.

Many of the trials included in the review were not of high quality, either due to poor or unclear allocation concealment or large losses to follow-up, which increases the risk of bias in the results. The data were also complicated by differing definitions of miscarriage. For some trials, miscarriage was considered to occur up until 26 or 28 weeks' gestation, while other studies reported miscarriage as pregnancy loss prior to 20 weeks' gestation, and stillbirth as pregnancy loss greater or equal to 20 weeks' gestation. Other studies did not specify their definition of miscarriage or stillbirth. In addition to the problems with differing definitions, the timing of the onset of vitamin supplementation for some of the included trials occurred in mid-pregnancy, which may limit the impact of supplementation on the risk of miscarriage. The review attempted to overcome these issues by using the outcome 'total fetal loss', which included either miscarriage or stillbirth.

In order to determine the effect of publication bias, we undertook funnel plots for comparisons with 10 or more studies. Overall, in the comparisons of any vitamins versus no vitamins, the funnel plots were symmetrical, suggesting a low possibility of publication bias. The graphs for the analysis of multivitamins were less symmetrical, indicating the possibility that smaller negative trials may be missing. However it is also possible that this asymmetry is due to differences in maternal characteristics between trials.

Women given any vitamins were also more likely to self-report abdominal pain in late pregnancy, although this outcome was only reported by one trial assessing vitamin C and E supplementation (Rumbold 2006). Few studies recorded or reported any information about side effects of vitamin supplementation. We are unaware of any other studies reporting an association between vitamins and abdominal pain; nevertheless, these findings highlight the need to assess potential side effects and adverse effects in trials.

No trials reported on any potential psychological effects such as anxiety and depression, for women experiencing miscarriage or stillbirth.

Authors' conclusions

Implications for practice

Any vitamins do not help prevent either early or late miscarriage or stillbirth. Supplementing women with vitamin A or multivitamins, with or without folic acid, may increase the risk of having a multiple birth, which may confer increases in perinatal morbidity and mortality. There is insufficient evidence to examine the effect of different combinations of vitamins on miscarriage, stillbirth and measures of infant growth.

Implications for research

The impact of different combinations of vitamins (i.e. multivitamin preparations with or without vitamin A and folic acid) on miscarriage, stillbirth, birthweight and measures of infant growth is unclear. Any future studies of vitamin supplementation should be high quality and focus on women at high risk of miscarriage, assess the most appropriate vitamin type and dosage, show it is beneficial without causing any harms to the mother or fetus and include assessments of any psychological effects and long-term follow-up of mothers and infants.

Acknowledgements

We thank Simon Gates for statistical advice regarding inclusion of cluster randomised trials, Lelia Duley for helpful comments on the format of the review and Sonja Henderson for assisting with review administration.

Data and analyses

Download statistical data

Comparison 1. Any vitamins versus no vitamins (or minimal vitamins)
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Total fetal loss (including miscarriages or combined miscarriages and stillbirths)1333943Relative risk (Fixed, 95% CI)1.04 [0.95, 1.14]
1.1 Trial entry before pregnancy525182Relative risk (Fixed, 95% CI)1.06 [0.96, 1.17]
1.2 Trial entry < 12 weeks' gestation1406Relative risk (Fixed, 95% CI)1.32 [0.63, 2.77]
1.3 Trial entry >= 12 weeks' and < 20 weeks' gestation1739Relative risk (Fixed, 95% CI)0.84 [0.38, 1.85]
1.4 Trial entry 'mixed' both < 20 and >= 20 weeks' gestation67616Relative risk (Fixed, 95% CI)0.89 [0.66, 1.20]
2 Early or late miscarriage1011266Risk Ratio (M-H, Fixed, 95% CI)1.09 [0.95, 1.25]
2.1 Trial entry before pregnancy47809Risk Ratio (M-H, Fixed, 95% CI)1.07 [0.93, 1.24]
2.2 Trial entry < 12 weeks' gestation1406Risk Ratio (M-H, Fixed, 95% CI)1.32 [0.63, 2.77]
2.3 Trial entry >= 12 weeks but < 20 weeks' gestation1739Risk Ratio (M-H, Fixed, 95% CI)1.32 [0.30, 5.87]
2.4 Trial entry 'mixed' both < 20 and >= 20 weeks' gestation42312Risk Ratio (M-H, Fixed, 95% CI)1.19 [0.63, 2.24]
3 Placental abruption5 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
3.1 Placental abruption44264Risk Ratio (M-H, Fixed, 95% CI)0.66 [0.34, 1.30]
3.2 Antepartum haemorrhage including placental abruption1200Risk Ratio (M-H, Fixed, 95% CI)7.0 [0.88, 55.86]
4 Psychological effects (anxiety and depression)00Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
5 Pre-eclampsia79561Risk Ratio (M-H, Random, 95% CI)0.88 [0.70, 1.09]
6 Stillbirth915980Risk Ratio (M-H, Fixed, 95% CI)0.86 [0.65, 1.13]
6.1 Trial entry before pregnancy37785Risk Ratio (M-H, Fixed, 95% CI)0.94 [0.48, 1.85]
6.2 Trial entry < 12 weeks' gestation00Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
6.3 Trial entry >= 12 weeks' but < 20 weeks' gestation1739Risk Ratio (M-H, Fixed, 95% CI)0.69 [0.27, 1.80]
6.4 Trial entry 'mixed' both < 20 and >= 20 weeks' gestation57456Risk Ratio (M-H, Fixed, 95% CI)0.86 [0.63, 1.19]
7 Perinatal death44313Risk Ratio (M-H, Fixed, 95% CI)0.83 [0.62, 1.11]
8 Neonatal death627657Relative risk (Fixed, 95% CI)1.11 [0.94, 1.31]
9 Preterm birth827414Relative risk (Fixed, 95% CI)1.02 [0.94, 1.10]
10 Very preterm birth44181Risk Ratio (M-H, Fixed, 95% CI)0.93 [0.75, 1.15]
11 Birthweight57497Mean Difference (IV, Random, 95% CI)16.99 [-37.66, 71.64]
12 Small-for-gestational age79356Risk Ratio (M-H, Fixed, 95% CI)0.96 [0.84, 1.08]
12.1 Birthweight less than 10th centile or birthweight < 2500 g79356Risk Ratio (M-H, Fixed, 95% CI)0.96 [0.84, 1.08]
13 Congenital malformations48933Risk Ratio (M-H, Fixed, 95% CI)1.47 [0.90, 2.40]
14 Multiple pregnancy320986Relative risk (Fixed, 95% CI)1.38 [1.12, 1.70]
15 Apgar score less than seven at five minutes1700Risk Ratio (M-H, Fixed, 95% CI)0.66 [0.27, 1.60]
16 Use of blood transfusion for the mother00Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
17 Anaemia (maternal)21190Risk Ratio (M-H, Random, 95% CI)0.90 [0.46, 1.73]
18 Anaemia (infant)1836Risk Ratio (M-H, Fixed, 95% CI)1.05 [0.98, 1.12]
19 Placental weight129Mean Difference (IV, Fixed, 95% CI)96.0 [30.73, 161.27]
20 Method of feeding14878Risk Ratio (M-H, Fixed, 95% CI)0.98 [0.96, 1.01]
20.1 Breastfeeding14878Risk Ratio (M-H, Fixed, 95% CI)0.98 [0.96, 1.01]
20.2 Formula00Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
20.3 Breastfeeding and formula00Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
21 Subsequent fertility00Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
22 Poor growth at childhood follow-up00Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
23 Disability at childhood follow-up00Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
24 Any adverse effects of vitamin supplementation sufficient to stop supplementation1739Risk Ratio (M-H, Fixed, 95% CI)1.16 [0.39, 3.41]
25 Maternal views of care00Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
26 Gynaecological hospital admission11365Risk Ratio (M-H, Fixed, 95% CI)0.20 [0.02, 1.69]
26.1 Any maternal admission to ICU11365Risk Ratio (M-H, Fixed, 95% CI)0.20 [0.02, 1.69]
27 Admission to neonatal intensive care unit2 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
27.1 Any admission to NICU11515Risk Ratio (M-H, Fixed, 95% CI)0.81 [0.59, 1.11]
27.2 > 4 days of NICU care11853Risk Ratio (M-H, Fixed, 95% CI)0.60 [0.27, 1.37]
27.3 > 7 days in NICU11515Risk Ratio (M-H, Fixed, 95% CI)0.87 [0.54, 1.39]
28 Healthcare costs00Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]
29 Duration of admission to the neonatal intensive care unit1181Mean Difference (IV, Fixed, 95% CI)1.30 [-0.28, 2.88]
30 Side effects11734Risk Ratio (M-H, Fixed, 95% CI)1.63 [1.12, 2.36]
30.1 Abdominal pain11734Risk Ratio (M-H, Fixed, 95% CI)1.63 [1.12, 2.36]
Analysis 1.1.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 1 Total fetal loss (including miscarriages or combined miscarriages and stillbirths).

Analysis 1.2.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 2 Early or late miscarriage.

Analysis 1.3.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 3 Placental abruption.

Analysis 1.5.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 5 Pre-eclampsia.

Analysis 1.6.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 6 Stillbirth.

Analysis 1.7.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 7 Perinatal death.

Analysis 1.8.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 8 Neonatal death.

Analysis 1.9.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 9 Preterm birth.

Analysis 1.10.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 10 Very preterm birth.

Analysis 1.11.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 11 Birthweight.

Analysis 1.12.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 12 Small-for-gestational age.

Analysis 1.13.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 13 Congenital malformations.

Analysis 1.14.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 14 Multiple pregnancy.

Analysis 1.15.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 15 Apgar score less than seven at five minutes.

Analysis 1.17.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 17 Anaemia (maternal).

Analysis 1.18.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 18 Anaemia (infant).

Analysis 1.19.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 19 Placental weight.

Analysis 1.20.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 20 Method of feeding.

Analysis 1.24.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 24 Any adverse effects of vitamin supplementation sufficient to stop supplementation.

Analysis 1.26.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 26 Gynaecological hospital admission.

Analysis 1.27.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 27 Admission to neonatal intensive care unit.

Analysis 1.29.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 29 Duration of admission to the neonatal intensive care unit.

Analysis 1.30.

Comparison 1 Any vitamins versus no vitamins (or minimal vitamins), Outcome 30 Side effects.

Comparison 2. Any vitamins (by quality)
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Total fetal loss (including miscarriage or combined miscarriages and stillbirths)5 Relative risk (Fixed, 95% CI)0.97 [0.84, 1.12]
1.1 Allocation concealment is adequate5 Relative risk (Fixed, 95% CI)0.97 [0.84, 1.12]
2 Early or late miscarriage44633Risk Ratio (M-H, Fixed, 95% CI)0.95 [0.71, 1.27]
2.1 Allocation concealment is adequate44633Risk Ratio (M-H, Fixed, 95% CI)0.95 [0.71, 1.27]
3 Stillbirth54916Risk Ratio (M-H, Fixed, 95% CI)0.85 [0.47, 1.53]
3.1 Allocation concealment is adequate54916Risk Ratio (M-H, Fixed, 95% CI)0.85 [0.47, 1.53]
Analysis 2.1.

Comparison 2 Any vitamins (by quality), Outcome 1 Total fetal loss (including miscarriage or combined miscarriages and stillbirths).

Analysis 2.2.

Comparison 2 Any vitamins (by quality), Outcome 2 Early or late miscarriage.

Analysis 2.3.

Comparison 2 Any vitamins (by quality), Outcome 3 Stillbirth.

Comparison 3. Vitamin C
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Total fetal loss6 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
1.1 Vitamin C + multivitamins versus placebo plus multivitamins1406Risk Ratio (M-H, Fixed, 95% CI)1.32 [0.63, 2.77]
1.2 Vitamin C and vitamin E versus placebo32899Risk Ratio (M-H, Fixed, 95% CI)0.82 [0.48, 1.42]
1.3 Vitamin C versus no supplement/placebo2224Risk Ratio (M-H, Fixed, 95% CI)1.28 [0.58, 2.83]
2 Early or late miscarriage5 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
2.1 Vitamin C + multivitamins versus placebo plus multivitamins1406Risk Ratio (M-H, Fixed, 95% CI)1.32 [0.63, 2.77]
2.2 Vitamin C and vitamin E versus placebo22616Risk Ratio (M-H, Fixed, 95% CI)0.70 [0.27, 1.84]
2.3 Vitamin C versus no supplement/placebo2224Risk Ratio (M-H, Fixed, 95% CI)1.17 [0.52, 2.65]
3 Antepartum haemorrhage and placental abruption5 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
3.1 Vitamin C and vitamin E versus placebo - placental abruption only44264Risk Ratio (M-H, Fixed, 95% CI)0.66 [0.34, 1.30]
3.2 Vitamin C versus placebo - antepartum haemorrhage including placental abruption1200Risk Ratio (M-H, Fixed, 95% CI)7.0 [0.88, 55.86]
4 Pre-eclampsia5 Risk Ratio (M-H, Random, 95% CI)Subtotals only
4.1 Vitamin C and vitamin E versus placebo44264Risk Ratio (M-H, Random, 95% CI)0.94 [0.72, 1.22]
4.2 Vitamin C versus placebo1200Risk Ratio (M-H, Random, 95% CI)1.0 [0.21, 4.84]
5 Stillbirth4 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
5.1 Vitamin C and vitamin E versus placebo32899Risk Ratio (M-H, Fixed, 95% CI)0.89 [0.46, 1.73]
5.2 Vitamin C versus placebo1200Risk Ratio (M-H, Fixed, 95% CI)3.0 [0.12, 72.77]
6 Perinatal death44313Risk Ratio (M-H, Fixed, 95% CI)0.83 [0.62, 1.11]
6.1 Vitamin C versus placebo1182Risk Ratio (M-H, Fixed, 95% CI)0.51 [0.05, 5.54]
6.2 Vitamin C and vitamin E versus placebo34131Risk Ratio (M-H, Fixed, 95% CI)0.84 [0.63, 1.12]
7 Neonatal death32717Risk Ratio (M-H, Fixed, 95% CI)0.69 [0.30, 1.61]
7.1 Vitamin C versus placebo1181Risk Ratio (M-H, Fixed, 95% CI)0.69 [0.12, 4.03]
7.2 Vitamin C and E versus placebo22536Risk Ratio (M-H, Fixed, 95% CI)0.69 [0.27, 1.81]
8 Preterm birth5 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
8.1 Vitamin C and vitamin E versus placebo44264Risk Ratio (M-H, Fixed, 95% CI)0.97 [0.85, 1.10]
8.2 Vitamin C versus placebo1200Risk Ratio (M-H, Fixed, 95% CI)1.43 [1.03, 1.99]
9 Very preterm birth44181Risk Ratio (M-H, Fixed, 95% CI)0.93 [0.75, 1.15]
9.1 Vitamin C versus placebo1200Risk Ratio (M-H, Fixed, 95% CI)1.3 [0.78, 2.17]
9.2 Vitamin C and vitamin E versus placebo33981Risk Ratio (M-H, Fixed, 95% CI)0.88 [0.70, 1.10]
10 Small-for-gestational age44233Risk Ratio (M-H, Fixed, 95% CI)0.90 [0.78, 1.04]
10.1 Vitamin C and vitamin E versus placebo44233Risk Ratio (M-H, Fixed, 95% CI)0.90 [0.78, 1.04]
11 Birthweight22561Mean Difference (IV, Fixed, 95% CI)1.25 [-47.45, 49.95]
11.1 Vitamin C and vitamin E versus placebo22561Mean Difference (IV, Fixed, 95% CI)1.25 [-47.45, 49.95]
12 Congenital malformations22254Risk Ratio (M-H, Fixed, 95% CI)1.44 [0.73, 2.84]
12.1 Vitamin C and vitamin E versus placebo22254Risk Ratio (M-H, Fixed, 95% CI)1.44 [0.73, 2.84]
13 Apgar score less than seven at five minutes1700Risk Ratio (M-H, Fixed, 95% CI)0.66 [0.27, 1.60]
13.1 Vitamin C and vitamin E versus placebo1700Risk Ratio (M-H, Fixed, 95% CI)0.66 [0.27, 1.60]
14 Any adverse effects of vitamin supplementation sufficient to stop supplementation1739Risk Ratio (M-H, Fixed, 95% CI)1.16 [0.39, 3.41]
14.1 Vitamin C and vitamin E versus placebo1739Risk Ratio (M-H, Fixed, 95% CI)1.16 [0.39, 3.41]
15 Gynaecological hospital admission11365Risk Ratio (M-H, Fixed, 95% CI)0.20 [0.02, 1.69]
15.1 Vitamin C and vitamin E versus placebo: Any maternal admission to ICU11365Risk Ratio (M-H, Fixed, 95% CI)0.20 [0.02, 1.69]
16 Admission to neonatal intensive care unit2 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
16.1 Vitamin C and vitamin E versus placebo: Any admission to NICU11515Risk Ratio (M-H, Fixed, 95% CI)0.81 [0.59, 1.11]
16.2 Vitamin C and vitamin E versus placebo: > 4 days of NICU care11853Risk Ratio (M-H, Fixed, 95% CI)0.60 [0.27, 1.37]
16.3 Vitamin C and vitamin E versus placebo: > 7 days of NICU care11515Risk Ratio (M-H, Fixed, 95% CI)0.87 [0.54, 1.39]
17 Side effects11734Risk Ratio (M-H, Fixed, 95% CI)1.63 [1.12, 2.36]
17.1 Vitamin C and vitamin E versus placebo: Abdominal pain11734Risk Ratio (M-H, Fixed, 95% CI)1.63 [1.12, 2.36]
Analysis 3.1.

Comparison 3 Vitamin C, Outcome 1 Total fetal loss.

Analysis 3.2.

Comparison 3 Vitamin C, Outcome 2 Early or late miscarriage.

Analysis 3.3.

Comparison 3 Vitamin C, Outcome 3 Antepartum haemorrhage and placental abruption.

Analysis 3.4.

Comparison 3 Vitamin C, Outcome 4 Pre-eclampsia.

Analysis 3.5.

Comparison 3 Vitamin C, Outcome 5 Stillbirth.

Analysis 3.6.

Comparison 3 Vitamin C, Outcome 6 Perinatal death.

Analysis 3.7.

Comparison 3 Vitamin C, Outcome 7 Neonatal death.

Analysis 3.8.

Comparison 3 Vitamin C, Outcome 8 Preterm birth.

Analysis 3.9.

Comparison 3 Vitamin C, Outcome 9 Very preterm birth.

Analysis 3.10.

Comparison 3 Vitamin C, Outcome 10 Small-for-gestational age.

Analysis 3.11.

Comparison 3 Vitamin C, Outcome 11 Birthweight.

Analysis 3.12.

Comparison 3 Vitamin C, Outcome 12 Congenital malformations.

Analysis 3.13.

Comparison 3 Vitamin C, Outcome 13 Apgar score less than seven at five minutes.

Analysis 3.14.

Comparison 3 Vitamin C, Outcome 14 Any adverse effects of vitamin supplementation sufficient to stop supplementation.

Analysis 3.15.

Comparison 3 Vitamin C, Outcome 15 Gynaecological hospital admission.

Analysis 3.16.

Comparison 3 Vitamin C, Outcome 16 Admission to neonatal intensive care unit.

Analysis 3.17.

Comparison 3 Vitamin C, Outcome 17 Side effects.

Comparison 4. Vitamin A
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Total fetal loss (including miscarriages or combined miscarriages and stillbirths)5 Relative risk (Fixed, 95% CI)Subtotals only
1.1 Vitamin A versus placebo111723Relative risk (Fixed, 95% CI)1.04 [0.92, 1.17]
1.2 B-carotene versus placebo111303Relative risk (Fixed, 95% CI)1.03 [0.91, 1.16]
1.3 Vitamin A versus B-carotene111720Relative risk (Fixed, 95% CI)1.01 [0.90, 1.14]
1.4 Vitamin A or B-carotene versus placebo117373Relative risk (Fixed, 95% CI)1.05 [0.91, 1.21]
1.5 Vitamin A (with/without multivitamins) versus multivitamins or placebo11074Relative risk (Fixed, 95% CI)0.80 [0.53, 1.21]
1.6 Vitamin A + iron + folate versus iron + folate31640Relative risk (Fixed, 95% CI)1.01 [0.61, 1.66]
2 Early or late miscarriage3 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
2.1 Vitamin A (with/without multivitamins) versus multivitamins or placebo11075Risk Ratio (M-H, Fixed, 95% CI)0.76 [0.37, 1.55]
2.2 Vitamin A + iron + folate versus iron + folate21397Risk Ratio (M-H, Fixed, 95% CI)0.87 [0.47, 1.63]
3 Stillbirth4 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
3.1 Vitamin A (with/without multivitamins) versus multivitamins or placebo11075Risk Ratio (M-H, Fixed, 95% CI)1.04 [0.60, 1.79]
3.2 Vitamin A + iron + folate versus iron + folate31640Risk Ratio (M-H, Fixed, 95% CI)1.29 [0.57, 2.91]
4 Neonatal death1 Relative risk (Fixed, 95% CI)Subtotals only
4.1 Vitamin A versus placebo110214Relative risk (Fixed, 95% CI)1.09 [0.92, 1.30]
4.2 B-carotene versus placebo19788Relative risk (Fixed, 95% CI)1.09 [0.91, 1.30]
4.3 Vitamin A versus B-carotene110228Relative risk (Fixed, 95% CI)1.0 [0.85, 1.18]
4.4 Vitamin A or B-carotene versus placebo115115Relative risk (Fixed, 95% CI)1.09 [0.91, 1.30]
5 Preterm birth3 Relative risk (Fixed, 95% CI)Subtotals only
5.1 Vitamin A versus placebo111723Relative risk (Fixed, 95% CI)1.04 [0.89, 1.21]
5.2 B-carotene versus placebo111303Relative risk (Fixed, 95% CI)1.01 [0.86, 1.18]
5.3 Vitamin A versus B-carotene111720Relative risk (Fixed, 95% CI)1.03 [0.88, 1.20]
5.4 Vitamin A or B-carotene versus placebo117373Relative risk (Fixed, 95% CI)1.02 [0.89, 1.17]
5.5 Vitamin A (with/without multivitamins) versus multivitamins or placebo11075Relative risk (Fixed, 95% CI)1.07 [0.84, 1.37]
5.6 Vitamin A + iron + folate versus iron + folate1700Relative risk (Fixed, 95% CI)1.11 [0.59, 2.09]
6 Birthweight1594Mean Difference (IV, Fixed, 95% CI)90.0 [2.68, 177.32]
6.1 Vitamin A + iron + folate versus iron + folate1594Mean Difference (IV, Fixed, 95% CI)90.0 [2.68, 177.32]
7 Small-for-gestational age11075Risk Ratio (M-H, Fixed, 95% CI)0.84 [0.58, 1.21]
7.1 Vitamin A (with/without multivitamins) versus multivitamins or placebo11075Risk Ratio (M-H, Fixed, 95% CI)0.84 [0.58, 1.21]
8 Multiple pregnancy1 Relative risk (Fixed, 95% CI)Subtotals only
8.1 Vitamin A versus placebo110697Relative risk (Fixed, 95% CI)1.35 [0.99, 1.85]
8.2 B-carotene versus placebo110294Relative risk (Fixed, 95% CI)1.37 [1.00, 1.88]
8.3 Vitamin A versus B-carotene110699Relative risk (Fixed, 95% CI)1.03 [0.77, 1.37]
8.4 Vitamin A or B-carotene versus placebo115845Relative risk (Fixed, 95% CI)1.39 [1.05, 1.84]
9 Very preterm birth11075Risk Ratio (M-H, Fixed, 95% CI)1.11 [0.71, 1.74]
9.1 Vitamin A (with/without multivitamins) versus multivitamins or placebo11075Risk Ratio (M-H, Fixed, 95% CI)1.11 [0.71, 1.74]
10 Maternal anaemia2 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
10.1 Vitamin A + beta-carotene with or without multivitamin versus placebo1807Risk Ratio (M-H, Fixed, 95% CI)0.86 [0.60, 1.24]
10.2 Vitamin A + beta-carotene versus placebo1539Risk Ratio (M-H, Fixed, 95% CI)0.91 [0.60, 1.38]
10.3 Vitamin A + iron and folic acid versus iron and folic acid1700Risk Ratio (M-H, Fixed, 95% CI)0.96 [0.82, 1.12]
11 Infant anaemia2 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
11.1 Infant anaemia at 6 weeks' of age - vitamin A + iron + folate versus iron + folate1562Risk Ratio (M-H, Fixed, 95% CI)0.58 [0.45, 0.75]
11.2 Infant anaemia at 12 months - vitamin A + iron + folate versus iron + folate1478Risk Ratio (M-H, Fixed, 95% CI)1.03 [0.88, 1.20]
11.3 Infant anaemia - vitamin A + beta-carotene with or without multivitamins versus placebo1625Risk Ratio (M-H, Fixed, 95% CI)0.99 [0.92, 1.06]
11.4 Infant anaemia - vitamin A + beta-carotene versus placebo1406Risk Ratio (M-H, Fixed, 95% CI)0.99 [0.92, 1.08]
12 Poor growth at childhood follow up2 Mean Difference (IV, Fixed, 95% CI)Subtotals only
12.1 Weight (g) at 6 weeks: vitamin A + iron + folate versus iron + folate1546Mean Difference (IV, Fixed, 95% CI)169.0 [16.55, 321.45]
12.2 Length (cm) at 6 weeks: vitamin A + iron + folate versus iron + folate1546Mean Difference (IV, Fixed, 95% CI)0.70 [0.15, 1.25]
12.3 Weight (g) at 4 months: vitamin A + iron + folate versus iron + folate1148Mean Difference (IV, Fixed, 95% CI)-100.0 [-377.14, 177.14]
12.4 Length (cm) at 4 months: vitamin A + iron + folate versus iron + folate1148Mean Difference (IV, Fixed, 95% CI)-0.5 [-1.33, 0.33]
Analysis 4.1.

Comparison 4 Vitamin A, Outcome 1 Total fetal loss (including miscarriages or combined miscarriages and stillbirths).

Analysis 4.2.

Comparison 4 Vitamin A, Outcome 2 Early or late miscarriage.

Analysis 4.3.

Comparison 4 Vitamin A, Outcome 3 Stillbirth.

Analysis 4.4.

Comparison 4 Vitamin A, Outcome 4 Neonatal death.

Analysis 4.5.

Comparison 4 Vitamin A, Outcome 5 Preterm birth.

Analysis 4.6.

Comparison 4 Vitamin A, Outcome 6 Birthweight.

Analysis 4.7.

Comparison 4 Vitamin A, Outcome 7 Small-for-gestational age.

Analysis 4.8.

Comparison 4 Vitamin A, Outcome 8 Multiple pregnancy.

Analysis 4.9.

Comparison 4 Vitamin A, Outcome 9 Very preterm birth.

Analysis 4.10.

Comparison 4 Vitamin A, Outcome 10 Maternal anaemia.

Analysis 4.11.

Comparison 4 Vitamin A, Outcome 11 Infant anaemia.

Analysis 4.12.

Comparison 4 Vitamin A, Outcome 12 Poor growth at childhood follow up.

Comparison 5. Multivitamin
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Total fetal loss (including miscarriages or combined miscarriages and stillbirths)12 Risk Ratio (M-H, Random, 95% CI)Subtotals only
1.1 Multivitamin + folic acid versus no multivitamin/folic acid36883Risk Ratio (M-H, Random, 95% CI)1.00 [0.75, 1.34]
1.2 Multivitamin without folic acid versus no multivitamin/folic acid1907Risk Ratio (M-H, Random, 95% CI)0.83 [0.56, 1.25]
1.3 Multivitamins with/without folic acid versus no multivitamins/folic acid11368Risk Ratio (M-H, Random, 95% CI)0.91 [0.65, 1.27]
1.4 Multivitamin + folic acid versus folic acid21096Risk Ratio (M-H, Random, 95% CI)1.03 [0.72, 1.48]
1.5 Multivitamin without folic acid versus folic acid21090Risk Ratio (M-H, Random, 95% CI)0.90 [0.62, 1.30]
1.6 Multivitamin with/without folic acid versus folic acid21644Risk Ratio (M-H, Random, 95% CI)0.95 [0.69, 1.30]
1.7 Multivitamin with/without vitamin A versus vitamin A or placebo11074Risk Ratio (M-H, Random, 95% CI)0.60 [0.39, 0.91]
1.8 Multivitamins versus control15021Risk Ratio (M-H, Random, 95% CI)0.83 [0.58, 1.17]
1.9 Multivitamin + vitamin E versus multivitamin without vitamin E or controls1823Risk Ratio (M-H, Random, 95% CI)0.92 [0.46, 1.83]
1.10 Multivitamins + iron + folic acid versus iron + folic acid542404Risk Ratio (M-H, Random, 95% CI)0.90 [0.75, 1.09]
2 Early or late miscarriage10 Risk Ratio (M-H, Random, 95% CI)Subtotals only
2.1 Multivitamin + folic acid versus no multivitamin/folic acid36883Risk Ratio (M-H, Random, 95% CI)0.99 [0.72, 1.38]
2.2 Multivitamin without folic acid versus no multivitamin/folic acid1907Risk Ratio (M-H, Random, 95% CI)0.89 [0.59, 1.34]
2.3 Multivitamin with/without folic acid versus no multivitamin/folic acid11368Risk Ratio (M-H, Random, 95% CI)0.95 [0.67, 1.34]
2.4 Multivitamin + folic acid versus folic acid21096Risk Ratio (M-H, Random, 95% CI)1.04 [0.72, 1.49]
2.5 Multivitamin without folic acid versus folic acid21090Risk Ratio (M-H, Random, 95% CI)0.89 [0.61, 1.31]
2.6 Multivitamin with/without folic acid versus folic acid21644Risk Ratio (M-H, Random, 95% CI)0.96 [0.70, 1.33]
2.7 Multivitamin + vitamin E versus multivitamin without vitamin E or controls1823Risk Ratio (M-H, Random, 95% CI)1.04 [0.26, 4.13]
2.8 Multivitamin + iron + folic acid versus iron + folic acid542404Risk Ratio (M-H, Random, 95% CI)0.90 [0.73, 1.11]
3 Placental abruption1 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
3.1 Multivitamins + iron + folic acid versus iron + folic acid160Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]
4 Pre-eclampsia2 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
4.1 Multivitamin versus control15021Risk Ratio (M-H, Fixed, 95% CI)0.70 [0.55, 0.90]
4.2 Multivitamin + iron + folic acid versus iron + folic acid160Risk Ratio (M-H, Fixed, 95% CI)0.24 [0.06, 1.01]
5 Stillbirth11 Risk Ratio (M-H, Random, 95% CI)Subtotals only
5.1 Multivitamin + folic acid versus no multivitamin/folic acid36883Risk Ratio (M-H, Random, 95% CI)1.04 [0.51, 2.10]
5.2 Multivitamin without folic acid versus no multivitamin/folic acid1907Risk Ratio (M-H, Random, 95% CI)0.14 [0.01, 2.76]
5.3 Multivitamin with/without folic acid versus no multivitamin/folic acid11368Risk Ratio (M-H, Random, 95% CI)0.33 [0.06, 1.97]
5.4 Multivitamin + folic acid versus folic acid21096Risk Ratio (M-H, Random, 95% CI)0.97 [0.14, 6.88]
5.5 Multivitamin without folic acid versus folic acid21090Risk Ratio (M-H, Random, 95% CI)0.99 [0.04, 22.88]
5.6 Multivitamin with/without folic acid versus folic acid21644Risk Ratio (M-H, Random, 95% CI)0.79 [0.15, 4.10]
5.7 Multivitamin versus control15021Risk Ratio (M-H, Random, 95% CI)0.83 [0.58, 1.17]
5.8 Multivitamin + vitamin E versus multivitamin without vitamin E or controls1823Risk Ratio (M-H, Random, 95% CI)0.88 [0.39, 1.98]
5.9 Multivitamin + iron + folic acid versus iron + folic acid542404Risk Ratio (M-H, Random, 95% CI)0.90 [0.75, 1.07]
6 Perinatal death5 Relative risk (Random, 95% CI)Subtotals only
6.1 Multivitamin + folic acid + iron + zinc + vitamin A versus folic acid + iron + zinc + vitamin A14308Relative risk (Random, 95% CI)1.11 [0.98, 1.26]
6.2 Multivitamin + iron + folic acid versus iron + folic acid442344Relative risk (Random, 95% CI)0.99 [0.80, 1.21]
7 Neonatal death8 Relative risk (Fixed, 95% CI)Subtotals only
7.1 Multivitamin + folic acid versus no multivitamin/folic acid14930Relative risk (Fixed, 95% CI)1.59 [0.30, 8.30]
7.2 Multivitamin + vitamin E versus multivitamin without vitamin E or controls1787Relative risk (Fixed, 95% CI)1.44 [0.91, 2.27]
7.3 Multivitamin versus control14895Relative risk (Fixed, 95% CI)1.0 [0.75, 1.34]
7.4 Multivitamin + folic acid + iron + zinc + vitamin A versus folic acid + iron + zinc + vitamin A14122Relative risk (Fixed, 95% CI)1.15 [0.97, 1.36]
7.5 Multivitamin + iron + folic acid versus iron + folic acid440706Relative risk (Fixed, 95% CI)0.91 [0.80, 1.03]
8 Preterm birth8 Relative risk (Fixed, 95% CI)Subtotals only
8.1 Multivitamin + folic acid versus no multivitamin/folic acid15502Relative risk (Fixed, 95% CI)1.01 [0.91, 1.12]
8.2 Multivitamin + folic acid + iron + zinc + vitamin A versus folic acid + iron + zinc + vitamin A13320Relative risk (Fixed, 95% CI)0.98 [0.90, 1.07]
8.3 Multivitamin + vitamin E versus multivitamin without vitamin E or controls1814Relative risk (Fixed, 95% CI)0.99 [0.85, 1.15]
8.4 Multivitamin + iron + folic acid versus iron + folic acid539540Relative risk (Fixed, 95% CI)1.00 [0.96, 1.04]
9 Very preterm birth18428Risk Ratio (M-H, Fixed, 95% CI)0.88 [0.73, 1.06]
9.1 Multivitamin + iron + folic acid versus iron + folic acid18428Risk Ratio (M-H, Fixed, 95% CI)0.88 [0.73, 1.06]
10 Birthweight4 Mean Difference (IV, Fixed, 95% CI)Subtotals only
10.1 Multivitamin + folic acid versus no multivitamin/folic acid14862Mean Difference (IV, Fixed, 95% CI)3.0 [-24.15, 30.15]
10.2 Multivitamin + iron + folic acid versus iron + folic acid310241Mean Difference (IV, Fixed, 95% CI)61.61 [37.32, 85.91]
11 Small-for-gestational age (birthweight less than the 10th percentile or < 2500 g8 Relative risk (Random, 95% CI)Subtotals only
11.1 Multivitamin + folic acid versus no multivitamin/folic acid14862Relative risk (Random, 95% CI)1.09 [0.94, 1.26]
11.2 Multivitamin + folic acid versus no multivitamin/folic acid (birthweight < 2500 g)1186Relative risk (Random, 95% CI)0.91 [0.63, 1.32]
11.3 Multivitamin + folic acid + iron + zinc + vitamin A versus folic acid + iron + zinc + vitamin A13320Relative risk (Random, 95% CI)0.98 [0.95, 1.02]
11.4 Multivitamin + folic acid + iron + zinc + vitamin A versus folic acid + iron + zinc + vitamin A (birthweight < 2500 g)13325Relative risk (Random, 95% CI)0.95 [0.90, 1.00]
11.5 Multivitamins + iron + folic acid versus iron + folic acid521434Relative risk (Random, 95% CI)0.93 [0.77, 1.12]
12 Congenital malformations4 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
12.1 Multivitamin + folic acid versus no multivitamin/folic acid25777Risk Ratio (M-H, Fixed, 95% CI)1.69 [0.81, 3.53]
12.2 Multivitamin without folic acid without versus no multivitamin/folic acid1907Risk Ratio (M-H, Fixed, 95% CI)1.60 [0.53, 4.86]
12.3 Multivitamin with/without folic acid versus no multivitamin/folic acid11368Risk Ratio (M-H, Fixed, 95% CI)1.99 [0.75, 5.26]
12.4 Multivitamin + folic acid versus folic acid21096Risk Ratio (M-H, Fixed, 95% CI)1.71 [0.72, 4.04]
12.5 Multivitamin without folic acid versus folic acid21090Risk Ratio (M-H, Fixed, 95% CI)1.61 [0.67, 3.85]
12.6 Multivitamin with/without folic acid versus folic acid21644Risk Ratio (M-H, Fixed, 95% CI)1.66 [0.76, 3.63]
12.7 Multivitamin + iron + folic acid versus iron + folic acid11200Risk Ratio (M-H, Fixed, 95% CI)1.0 [0.14, 7.08]
13 Multiple pregnancy25141Risk Ratio (M-H, Fixed, 95% CI)1.36 [1.00, 1.85]
13.1 Multivitamin + folic acid versus no multivitamin/folic acid25141Risk Ratio (M-H, Fixed, 95% CI)1.36 [1.00, 1.85]
14 Maternal anaemia4 Relative risk (Fixed, 95% CI)Subtotals only
14.1 Multivitamin + folic acid + iron + zinc + vitamin A versus folic acid + iron + zinc + vitamin A (any anaemia)1813Relative risk (Fixed, 95% CI)0.92 [0.83, 1.03]
14.2 Multivitamin + folic acid + iron + zinc+vitamin A versus folic acid + iron + zinc + vitamin A (severe anaemia)1813Relative risk (Fixed, 95% CI)0.82 [0.53, 1.27]
14.3 Multivitamins versus placebo1538Relative risk (Fixed, 95% CI)0.78 [0.50, 1.22]
14.4 Multivitamins + vitamin A + beta-carotene versus placebo1535Relative risk (Fixed, 95% CI)0.82 [0.53, 1.26]
14.5 Multivitamins + iron + folic acid versus iron + folic acid22278Relative risk (Fixed, 95% CI)0.88 [0.81, 0.96]
15 Breastfeeding14878Risk Ratio (M-H, Fixed, 95% CI)0.98 [0.96, 1.01]
15.1 Multivitamin versus control14878Risk Ratio (M-H, Fixed, 95% CI)0.98 [0.96, 1.01]
16 Poor growth at childhood follow up: Underweight in childhood (6-8 years of age)13356Risk Ratio (Fixed, 95% CI)1.05 [0.97, 1.13]
16.1 Multivitamin + folic acid + iron + zinc + vitamin A versus folic acid + iron + zinc + vitamin A13356Risk Ratio (Fixed, 95% CI)1.05 [0.97, 1.13]
17 Poor growth at childhood follow up: Stunting in childhood (6-8 years of age)13356Risk Ratio (Fixed, 95% CI)1.09 [1.00, 1.19]
17.1 Multivitamin + folic acid + iron + zinc + vitamin A versus folic acid + iron + zinc + vitamin A13356Risk Ratio (Fixed, 95% CI)1.09 [1.00, 1.19]
18 Additional outcomes - infant death14122Relative risk (Fixed, 95% CI)1.10 [0.94, 1.29]
18.1 Multivitamin + folic acid + iron + zinc + vitamin A versus folic acid + iron + zinc + vitamin A14122Relative risk (Fixed, 95% CI)1.10 [0.94, 1.29]
Analysis 5.1.

Comparison 5 Multivitamin, Outcome 1 Total fetal loss (including miscarriages or combined miscarriages and stillbirths).

Analysis 5.2.

Comparison 5 Multivitamin, Outcome 2 Early or late miscarriage.

Analysis 5.3.

Comparison 5 Multivitamin, Outcome 3 Placental abruption.

Analysis 5.4.

Comparison 5 Multivitamin, Outcome 4 Pre-eclampsia.

Analysis 5.5.

Comparison 5 Multivitamin, Outcome 5 Stillbirth.

Analysis 5.6.

Comparison 5 Multivitamin, Outcome 6 Perinatal death.

Analysis 5.7.

Comparison 5 Multivitamin, Outcome 7 Neonatal death.

Analysis 5.8.

Comparison 5 Multivitamin, Outcome 8 Preterm birth.

Analysis 5.9.

Comparison 5 Multivitamin, Outcome 9 Very preterm birth.

Analysis 5.10.

Comparison 5 Multivitamin, Outcome 10 Birthweight.

Analysis 5.11.

Comparison 5 Multivitamin, Outcome 11 Small-for-gestational age (birthweight less than the 10th percentile or < 2500 g.

Analysis 5.12.

Comparison 5 Multivitamin, Outcome 12 Congenital malformations.

Analysis 5.13.

Comparison 5 Multivitamin, Outcome 13 Multiple pregnancy.

Analysis 5.14.

Comparison 5 Multivitamin, Outcome 14 Maternal anaemia.

Analysis 5.15.

Comparison 5 Multivitamin, Outcome 15 Breastfeeding.

Analysis 5.16.

Comparison 5 Multivitamin, Outcome 16 Poor growth at childhood follow up: Underweight in childhood (6-8 years of age).

Analysis 5.17.

Comparison 5 Multivitamin, Outcome 17 Poor growth at childhood follow up: Stunting in childhood (6-8 years of age).

Analysis 5.18.

Comparison 5 Multivitamin, Outcome 18 Additional outcomes - infant death.

Comparison 6. Folic acid
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Total fetal loss (including miscarriages or combined miscarriages and stillbirths)6 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
1.1 Folic acid + multivitamin versus no folic acid/multivitamin36883Risk Ratio (M-H, Fixed, 95% CI)1.09 [0.95, 1.25]
1.2 Folic acid without multivitamin versus no folic acid/multivitamin1903Risk Ratio (M-H, Fixed, 95% CI)0.95 [0.64, 1.40]
1.3 Folic acid with/without multivitamin versus no folic acid/multivitamin11364Risk Ratio (M-H, Fixed, 95% CI)0.97 [0.69, 1.35]
1.4 Folic acid + multivitamin versus multivitamin21102Risk Ratio (M-H, Fixed, 95% CI)1.15 [0.80, 1.67]
1.5 Folic acid without multivitamin versus multivitamin21090Risk Ratio (M-H, Fixed, 95% CI)1.12 [0.77, 1.62]
1.6 Folic acid with or without multivitamin versus multivitamin21644Risk Ratio (M-H, Fixed, 95% CI)1.14 [0.82, 1.57]
1.7 Folic acid + iron versus iron175Risk Ratio (M-H, Fixed, 95% CI)0.23 [0.01, 4.59]
1.8 Folic acid + iron + antimalarials versus iron + antimalarials1160Risk Ratio (M-H, Fixed, 95% CI)13.0 [0.74, 226.98]
2 Early or late miscarriage6 Risk Ratio (M-H, Random, 95% CI)Subtotals only
2.1 Folic acid + multivitamin versus no folic acid/multivitamin36883Risk Ratio (M-H, Random, 95% CI)0.99 [0.72, 1.38]
2.2 Folic acid without multivitamins versus no folic acid/multivitamin1903Risk Ratio (M-H, Random, 95% CI)0.97 [0.65, 1.44]
2.3 Folic acid with/without multivitamin versus no folic acid/multivitamin11364Risk Ratio (M-H, Random, 95% CI)0.99 [0.70, 1.39]
2.4 Folic acid + multivitamin versus multivitamin21102Risk Ratio (M-H, Random, 95% CI)1.16 [0.80, 1.69]
2.5 Folic acid without multivitamin versus multivitamin21090Risk Ratio (M-H, Random, 95% CI)1.12 [0.77, 1.64]
2.6 Folic acid with/without multivitamin versus multivitamin21642Risk Ratio (M-H, Random, 95% CI)1.09 [0.79, 1.51]
2.7 Folic acid + iron versus iron175Risk Ratio (M-H, Random, 95% CI)0.38 [0.02, 9.03]
2.8 Folic acid + iron + antimalarials versus iron + antimalarials1160Risk Ratio (M-H, Random, 95% CI)13.0 [0.74, 226.98]
3 Pre-eclampsia175Risk Ratio (M-H, Fixed, 95% CI)1.14 [0.17, 7.69]
3.1 Folic acid + iron versus iron175Risk Ratio (M-H, Fixed, 95% CI)1.14 [0.17, 7.69]
4 Stillbirth5 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
4.1 Folic acid + multivitamin versus no folic acid/multivitamin36883Risk Ratio (M-H, Fixed, 95% CI)1.03 [0.51, 2.09]
4.2 Folic acid without multivitamin versus no folic acid/multivitamin1903Risk Ratio (M-H, Fixed, 95% CI)0.67 [0.11, 4.02]
4.3 Folic acid with/without multivitamin versus no folic acid/multivitamin11364Risk Ratio (M-H, Fixed, 95% CI)0.67 [0.15, 2.96]
4.4 Folic acid + multivitamin versus multivitamin21102Risk Ratio (M-H, Fixed, 95% CI)1.00 [0.20, 4.99]
4.5 Folic acid without multivitamin versus multivitamin21090Risk Ratio (M-H, Fixed, 95% CI)4.97 [0.58, 42.29]
4.6 Folic acid with/without multivitamin versus multivitamin21644Risk Ratio (M-H, Fixed, 95% CI)0.84 [0.20, 3.53]
4.7 Folic acid + iron versus iron175Risk Ratio (M-H, Fixed, 95% CI)0.38 [0.02, 9.03]
5 Perinatal death14308Relative risk (Fixed, 95% CI)0.97 [0.85, 1.11]
5.1 Folic acid + iron + zinc + multivitamin + vitamin A versus vitamin A14308Relative risk (Fixed, 95% CI)0.97 [0.85, 1.11]
6 Neonatal death2 Relative risk (Fixed, 95% CI)Subtotals only
6.1 Folic acid + multivitamin versus no folic acid/multivitamin14930Relative risk (Fixed, 95% CI)1.59 [0.30, 8.28]
6.2 Folic acid + iron + zinc + multivitamin + vitamin A versus vitamin A14122Relative risk (Fixed, 95% CI)0.96 [0.80, 1.14]
7 Preterm birth3 Relative risk (Fixed, 95% CI)Subtotals only
7.1 Folic acid + multivitamin versus no folic acid/multivitamin15502Relative risk (Fixed, 95% CI)1.01 [0.91, 1.12]
7.2 Folic acid + multivitamin versus no folic acid/multivitamin175Relative risk (Fixed, 95% CI)1.01 [0.65, 1.56]
7.3 Folic acid + iron + zinc + multivitamin + vitamin A versus vitamin A13320Relative risk (Fixed, 95% CI)1.02 [0.94, 1.11]
8 Birthweight3 Mean Difference (IV, Fixed, 95% CI)Subtotals only
8.1 Folic acid + multivitamin versus no folic acid/multivitamin14862Mean Difference (IV, Fixed, 95% CI)3.0 [-24.15, 30.15]
8.2 Folic acid versus placebo129Mean Difference (IV, Fixed, 95% CI)312.0 [108.52, 515.48]
8.3 Folic + iron versus control145Mean Difference (IV, Fixed, 95% CI)-32.0 [-213.62, 149.62]
9 Small-for-gestational age4 Relative risk (Fixed, 95% CI)Subtotals only
9.1 Folic acid + multivitamin versus no folic acid/multivitamin14862Relative risk (Fixed, 95% CI)1.09 [0.94, 1.26]
9.2 Folic acid + iron + zinc + multivitamin + vitamin A versus vitamin A13320Relative risk (Fixed, 95% CI)0.97 [0.92, 1.03]
9.3 Folic acid + iron versus iron (birthweight < 2500 g)175Relative risk (Fixed, 95% CI)1.06 [0.48, 2.33]
9.4 Folic acid + multivitamin versus no folic acid/multivitamin (birthweight < 2500 g)1186Relative risk (Fixed, 95% CI)0.91 [0.63, 1.32]
9.5 Folic acid + iron + zinc + multivitamin + vitamin A versus vitamin A (birthweight < 2500 g)13325Relative risk (Fixed, 95% CI)0.94 [0.90, 0.99]
10 Congenital malformations3 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
10.1 Folic acid + multivitamin versus no folic acid/multivitamin25777Risk Ratio (M-H, Fixed, 95% CI)1.69 [0.81, 3.53]
10.2 Folic acid without multivitamin versus no folic acid/multivitamin1903Risk Ratio (M-H, Fixed, 95% CI)1.42 [0.45, 4.43]
10.3 Folic acid with/without multivitamin versus no folic acid/multivitamin11364Risk Ratio (M-H, Fixed, 95% CI)1.90 [0.71, 5.04]
10.4 Folic acid + multivitamin versus multivitamin21102Risk Ratio (M-H, Fixed, 95% CI)1.07 [0.51, 2.26]
10.5 Folic acid without multivitamin versus multivitamin21090Risk Ratio (M-H, Fixed, 95% CI)0.62 [0.26, 1.49]
10.6 Folic acid with or without multvitamin versus multivitamin21644Risk Ratio (M-H, Fixed, 95% CI)0.85 [0.43, 1.67]
11 Multiple pregnancy2 Risk Ratio (M-H, Fixed, 95% CI)Subtotals only
11.1 Folic acid + multivitamin versus no folic acid/multivitamin25141Risk Ratio (M-H, Fixed, 95% CI)1.36 [1.00, 1.85]
12 Maternal anaemia3 Relative risk (Fixed, 95% CI)Subtotals only
12.1 Folic acid + iron + zinc + multivitamin + vitamin A versus vitamin A (any anaemia)1813Relative risk (Fixed, 95% CI)0.83 [0.77, 0.91]
12.2 Folic acid + iron + zinc + multivitamin + vitamin A versus vitamin A (severe anaemia)1813Relative risk (Fixed, 95% CI)0.82 [0.59, 1.16]
12.3 Folic acid + iron versus iron (severe anaemia)185Relative risk (Fixed, 95% CI)1.06 [0.25, 4.42]
12.4 Folic acid + iron versus no folic acid or iron189Relative risk (Fixed, 95% CI)1.53 [0.79, 2.95]
13 Poor growth in childhood: Stunting in childhood (6-8 years of age)13356Risk Ratio (Fixed, 95% CI)0.93 [0.86, 1.00]
13.1 Folic acid + iron + zinc + vitamin A versus multivitamin + vitamin A13356Risk Ratio (Fixed, 95% CI)0.93 [0.86, 1.00]
14 Poor growth in childhood: Underweight in childhood (6-8 years of age)13356Risk Ratio (Fixed, 95% CI)0.97 [0.91, 1.04]
14.1 Folic acid + iron + zinc + vitamin A versus multivitamin + vitamin A13356Risk Ratio (Fixed, 95% CI)0.97 [0.91, 1.04]
15 Placental weight129Mean Difference (IV, Fixed, 95% CI)96.0 [30.73, 161.27]
15.1 Folic acid versus placebo129Mean Difference (IV, Fixed, 95% CI)96.0 [30.73, 161.27]
16 Additional outcomes - infant death14122Relative risk (Fixed, 95% CI)0.95 [0.81, 1.11]
16.1 Folic acid + iron + zinc + multivitamin + vitamin A versus vitamin A14122Relative risk (Fixed, 95% CI)0.95 [0.81, 1.11]
Analysis 6.1.

Comparison 6 Folic acid, Outcome 1 Total fetal loss (including miscarriages or combined miscarriages and stillbirths).

Analysis 6.2.

Comparison 6 Folic acid, Outcome 2 Early or late miscarriage.

Analysis 6.3.

Comparison 6 Folic acid, Outcome 3 Pre-eclampsia.

Analysis 6.4.

Comparison 6 Folic acid, Outcome 4 Stillbirth.

Analysis 6.5.

Comparison 6 Folic acid, Outcome 5 Perinatal death.

Analysis 6.6.

Comparison 6 Folic acid, Outcome 6 Neonatal death.

Analysis 6.7.

Comparison 6 Folic acid, Outcome 7 Preterm birth.

Analysis 6.8.

Comparison 6 Folic acid, Outcome 8 Birthweight.

Analysis 6.9.

Comparison 6 Folic acid, Outcome 9 Small-for-gestational age.

Analysis 6.10.

Comparison 6 Folic acid, Outcome 10 Congenital malformations.

Analysis 6.11.

Comparison 6 Folic acid, Outcome 11 Multiple pregnancy.

Analysis 6.12.

Comparison 6 Folic acid, Outcome 12 Maternal anaemia.

Analysis 6.13.

Comparison 6 Folic acid, Outcome 13 Poor growth in childhood: Stunting in childhood (6-8 years of age).

Analysis 6.14.

Comparison 6 Folic acid, Outcome 14 Poor growth in childhood: Underweight in childhood (6-8 years of age).

Analysis 6.15.

Comparison 6 Folic acid, Outcome 15 Placental weight.

Analysis 6.16.

Comparison 6 Folic acid, Outcome 16 Additional outcomes - infant death.

Appendices

Appendix 1. Additional searching carried out for the initial version of the review

For the initial version of the review, authors carried out a separate search of CENTRAL (The Cochrane Library, 2003, Issue 2) for the following terms: miscarriage*, spontaneous abortion, recurrent abortion, spontaneous pregnancy loss, recurrent pregnancy loss, fetal death, vitamin*, folate, folic acid; and also MEDLINE (1966 to May 2003), Current Contents (1998 to May 2003) and EMBASE (1980 to May 2003) using the search strategy given below:

  1. miscarriage*

  2. spontaneous abortion

  3. recurrent abortion

  4. habitual abortion

  5. spontaneous pregnancy loss

  6. recurrent pregnancy loss

  7. early pregnancy loss

  8. early pregnancy bleeding

  9. fetal death

  10. #1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9

  11. vitamin*

  12. retinol*

  13. carotenoid*

  14. thiamin*

  15. riboflavin

  16. niacin or nicotinamide or nicotinic acid

  17. pantothenic acid or pantothenate

  18. pyridox*

  19. cyanocobalamin or cobalamin

  20. ascorb*

  21. calciferol

  22. tocopherol* or alpha-tocopherol

  23. folate*

  24. folic acid

  25. phylloquinone

  26. menaquinone

  27. #13 or #14 or #15 or #16 or #17 or #18 or #19 or #20 or #21 or #22 or #23 or #24 or #25 or #26 or #11 or #12

  28. #10 and #27

  29. random*

  30. controlled-clinical-trial

  31. #29 or #30

  32. #28 and #31

What's new

DateEventDescription
27 August 2010New search has been performedSearch updated. 11 new studied included (Fawzi 2007; Fleming 1986; Osrin 2005; Roberfroid 2008; Rumbold 2006; Rumiris 2006; Spinnato 2007; Taylor 1982; The Summit 2008; Van den Broek 2006; Villar 2009), 3 studies excluded (Feyi-Waboso 2005; Huybregts 2009; Shu 2002). Two new studies are awaiting classification (Chelchowska 2004; Kubik 2004) and two new ongoing trials were identified (Fall 2007; Johns 2004; Sezikawa 2007).
27 August 2010New citation required but conclusions have not changedSubstantive amendment and addition of a new author.

History

Protocol first published: Issue 1, 2003
Review first published: Issue 2, 2005

DateEventDescription
20 September 2008AmendedConverted to new review format.

Contributions of authors

Alice Rumbold developed and wrote the protocol, extracted data and prepared the review. Ning Pan and Philippa Middleton extracted data and were involved in the analysis and writing of the review. Caroline Crowther commented on and revised the various drafts of the protocol and review during its development.

Declarations of interest

Alice Rumbold and Caroline Crowther are investigators on the Australian Collaborative Trial of Supplements with vitamin C and vitamin E for the prevention of pre-eclampsia (Rumbold 2006). This trial is included in this review but its eligibility for inclusion, trial quality assessments and data extraction were carried out independently by two of the review authors not involved in the original trial.

Sources of support

Internal sources

  • Department of Obstetrics and Gynaecology, The University of Adelaide, Australia.

  • The University of Adelaide Medical Endowment Fund, Australia.

    Dr Rumbold is supported by the Jean B Reid Fellowship

External sources

  • Department of Health and Ageing, Australia.

  • Department of Nutrition for Health and Development, World Health Organization, Switzerland.

    Provided funding for the preparation of this updated review.

Differences between protocol and review

We now include trials where supplementation occurred in mid-pregnancy. This was not specified in the original protocol for this review, but this was amended to be in line with other miscarriage reviews such as 'Progestogen for preventing miscarriage' (Haas 2009). We included trials where the onset of supplementation occurred both prior to and after 20 weeks' gestation, and when it could not be established whether the majority of the women started supplementation prior to 20 weeks' gestation. To overcome differences in the definition of miscarriage and stillbirth, we have used a combined outcome of total fetal loss (early or late miscarriage or stillbirth). We have still reported early or late miscarriage and stillbirth separately in addition to this combined outcome. Similarly, we specified in the original protocol that we would exclude studies reporting greater than 20% losses to follow-up. In this review we have included studies that reported more than 20% losses to follow-up and undertaken further analyses based on trial quality.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Briscoe 1959

Methods

Randomisation and allocation concealment: unclear, no methodological details given, dubious as the number of women allocated to the treatment group was more than double that allocated to the placebo group. "Unselected patients were each given 200 capsules... these were given a code, unknown to us and contained either an inert powder or 100 mg each of ascorbic acid and hesperidin."

Blinding of outcome assessment: women and study investigators did not know the treatment codes.

Documentation of exclusion: none reported.

Use of placebo control: placebo given; however, all women received an additional multivitamin supplement.

Participants406 women were recruited in the study. Eligible women were "unselected patients" in private obstetrics care, that were less than or equal to 10 weeks' pregnant, and were eligible regardless of whether they were currently bleeding or the number of previous pregnancies. Women greater than 10 weeks' gestation were excluded. 406 women were randomised to either vitamin C (n = 303) or placebo (n = 103), no losses to follow-up were reported. 77 women in the study had more than 2 previous miscarriages and/or bleeding in the pregnancy, and 329 had 2 or fewer miscarriages and no bleeding in the pregnancy.
InterventionsAll women were given 200 tablets, containing either 100 mg each of ascorbic acid and hesperidin or placebo (an inert powder).
The study lasted for 7 weeks. For the first two weeks, women were asked to take 8 tablets daily (i.e. daily 800 mg each of vitamin C and hesperidin or placebo). For the following 5 weeks, women took 4 tablets daily (i.e. daily 400 mg each of vitamin C and hesperidin or placebo). All women received a multiple vitamin supplement containing 50 mg vitamin C.
Outcomes
  1. Spontaneous miscarriage.

  2. Spontaneous miscarriage in women with 2 or fewer previous miscarriages and no bleeding in the current pregnancy.

  3. Spontaneous miscarriage in women with more than 2 previous miscarriages and/or bleeding in the current pregnancy.

  4. Spontaneous miscarriage in women who experienced recurrent miscarriage.

NotesWomen's risk of spontaneous and recurrent miscarriage is unclear, as there is no information about concurrent medical conditions or other risk factors for miscarriage. 9 of the 406 women were classified as experiencing recurrent miscarriage.
No information is available about women's nutritional status.
No sample-size calculation reported.
Intention-to-treat analyses performed (no losses to follow-up reported).
Compliance: no compliance information reported.
Location: Philadelphia, USA.
Timeframe: unclear.
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Unclear riskNo methodological details given.
Allocation concealment?Unclear riskNo methodological details given.
Blinding?
All outcomes
Low riskWomen and study investigators did not know the treatment allocation.
Incomplete outcome data addressed?
All outcomes
Low riskNo losses to follow up reported.
Free of selective reporting?Unclear riskLimited information about selection bias, stated that 'unselected patients' were included.
Free of other bias?Unclear riskLimited methodological details provided including patient compliance.

Chappell 1999

Methods

Randomisation and allocation concealment: a computer-generated randomisation list using blocks of 10 was given to the hospital pharmacy departments. Researchers allocated the next available number to participants and women collected the trial tablets from the pharmacy department.

Blinding of outcome assessment: women, caregivers and researchers were blinded to the treatment allocation until recruitment, data collection and laboratory analyses were complete.

Documentation of exclusion: 123 (43.5%) women were excluded, of which 70 women were withdrawn because their second Doppler scan was normal. Pregnancy outcome data were reported for all women randomised.

Use of placebo control: placebo control.

Participants283 women were recruited into the study. Inclusion criteria: abnormal Doppler waveform in either uterine artery at 18-22 weeks' gestation or a history in the preceding pregnancy of pre-eclampsia necessitating delivery before 37 weeks' gestation, eclampsia or the syndrome of HELLP.
Exclusion criteria: heparin or warfarin treatment, abnormal fetal-anomaly scan or multiple pregnancy.
Women were randomised at 18-22 weeks' gestation; however, women with a previous history who were identified at an earlier stage were randomised at 16 weeks' gestation. Women with abnormal Doppler waveform analysis returned for a second scan at 24 weeks' gestation, those with a normal waveform at this time stopped treatment and were withdrawn from the study. The remaining women who had persistently abnormal waveforms, and those with a previous history or pre-eclampsia remained in the study and were seen every 4 weeks through the rest of pregnancy. 1512 women underwent Doppler screening, 273 women had abnormal waveforms and of these, 242 women consented to the study. An additional 41 women who had a history of pre-eclampsia consented. 283 women were randomised to either the vitamin C and E group (n = 141) or the placebo group (n = 142), 72 women had normal Doppler scans at 24 weeks' gestation and 24 women did not return for a second scan and were withdrawn. A further 27 women withdrew from the trial after 24 weeks' gestation for various reasons. In total, 160 women completed the trial protocol until delivery, 79 in the vitamin C and E group and 81 in the placebo group. Pregnancy outcome data were presented for all women randomised (n = 283) as well as only for those women completing the trial protocol (n = 160).
InterventionsWomen randomised to the vitamin C and E group received tablets containing 1000 mg vitamin C daily and capsules containing 400 IU vitamin E daily.
Women randomised to the placebo group received tablets containing microcrystalline cellulose and soya bean oil, that were identical in appearance to the vitamin C tablets and vitamin E capsules. After 24 weeks' gestation women were seen every 4 weeks, and blood samples were taken at each visit.
Outcomes
  1. Ratio of PAI-1 to PAI-2.

  2. Incidence of pre-eclampsia.

  3. Placental abruption.

  4. Spontaneous preterm delivery (< 37 weeks).

  5. Intrauterine death.

  6. Small-for-gestational-age infants (on or below the 10th centile).

  7. Mean systolic and diastolic blood pressure before delivery.

  8. Gestational age at delivery (median, IQR).

  9. Birthweight (median, IQR).

  10. Birthweight centile (median, IQR).

NotesWomen's risk of spontaneous and recurrent miscarriage is unclear, women were at high risk of pre-eclampsia.
No information is available about women's nutritional status.
Sample-size calculation reported, based on a 30% reduction in PAI-1.
Intention-to-treat analyses performed.
Compliance: "within the treated group, plasma ascorbic acid concentration increased by 32% from baseline values and plasma alpha-tocopherol increased by 54%".
Location: London, UK.
Timeframe: unclear.
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Low riskComputer generated random number list.
Allocation concealment?Low riskRandom number list used blocks of 10 and was held by the pharmacy department.
Blinding?
All outcomes
Low riskWomen, caregivers and researchers were blinded until the analyses were completed.
Incomplete outcome data addressed?
All outcomes
Low risk123 (43.5%) women were excluded, of which, 70 women were withdrawn because their second Doppler scan was normal. Data were reported for all women randomised.
Free of selective reporting?Low riskData reported for all outcomes in methods.
Free of other bias?Low riskThe study appears to be free of other sources of bias.

Christian 2003

Methods

Randomisation and allocation concealment: cluster randomisation of 30 "village development communities" using blocks of 5 within each community, randomisation occurred by "drawing numbered identical chits from a hat".

Blinding of outcome assessment: women, field staff, investigators and statisticians did not know the treatment codes until the end of the study.

Documentation of exclusion: 534 (10.7%) women or infants were excluded and 343 (6.7%) infants were lost to follow up.

Use of placebo control: no placebo given, women in the control group were given vitamin A only.

ParticipantsAll women of reproductive age in the 30 village development communities were considered eligible. Women who were currently pregnant, breastfeeding a baby < 9 months old, menopausal, sterilised or widowed were excluded.
Within the timeframe, 14,185 women were identified as likely to become pregnant. Of these, 4998 pregnancies were confirmed with urine testing; however, 4926 women remained in the trial with 72 women excluded either due to false positive pregnancy testing, unknown outcomes or induced abortions. Women were allocated to either vitamin A control (n = 1037), folic acid (n = 929), folic acid-iron (n = 940), folic acid-iron-zinc (n = 982) or multiple micronutrients (n = 1038). 830 pregnancies (16.8%) ended in either miscarriage, stillbirth or maternal death. The remaining pregnancies resulted in 4130 livebirths. Of these, 805 (19.5%) were excluded as they were either lost to follow-up or birthweight was measured after 72 hours after birth. The final analysis involved 3325 infants allocated to control (n = 685), folic acid (n = 628), folic acid-iron (n = 635), folic acid-iron-zinc (n = 672) or multiple micronutrients (n = 705).
Interventions

Women were allocated to one of five groups:

  1. control (1000 mcg vitamin A);

  2. folic acid (400 mcg, 1000 mcg vitamin A);

  3. folic acid-iron (60 mg ferrous fumarate, 400 mcg folic acid, 1000 mcg vitamin A);

  4. folic acid-iron-zinc (30 mg zinc sulphate, 60 mg ferrous fumarate, 400 mcg folic acid, 1000 mcg vitamin A);

  5. multiple micronutrients-folic acid-iron-zinc (60 mg ferrous fumarate, 400 mcg folic acid, 30 mg zinc sulphate,1000 mcg vitamin A, 10 mcg vitamin D,10 mg vitamin E, 1.6 mg vitamin B-1, 1.8 mg vitamin B-2, 20 mg niacin, 2.2 mg vitamin B-6, 2.6 mcg vitamin B12, 100 mg vitamin C, 65 mcg vitamin K, 2.0 mg copper, 100 mg magnesium).

At enrolment women received 15 caplets and were instructed to take one caplet every night. Women were then visited by field staff twice a week to monitor compliance and replenish supplies of the caplets.

Outcomes
  1. Perinatal death, defined as stillbirths (gestational age >= 28 wk) and deaths among liveborn infants in the first 7 days of life.

  2. Neonatal deaths, defined as deaths from 0 to 28 days of life.

  3. Infant death, defined as deaths from 0 to 90 days of life.

  4. Birthweight.

  5. Length.

  6. Chest circumference.

  7. Head circumference.

  8. Low birthweight (< 2500 g).

  9. Small-for-gestational age (below 10th centile for USA national reference for fetal growth).

  10. Preterm birth (< 37 weeks).

"The rate if miscarriage did not differ by treatment group and ranged between 12% and 15% (data not show)". "Miscarriage was defined as a pregnancy that ended in a fetal loss before 28 wk of gestation."

NotesThe following information was given about multiple births: "the numbers of twin pregnancies (34 pairs of liveborn twins and 8 pairs with one stillborn) was comparable across treatment groups".
Women's risk of spontaneous and recurrent miscarriage is unclear, as there is no information about concurrent medical conditions or other risk factors for miscarriage.
Information on women's diet was recorded; however, no information was reported about micronutrient intake, including vitamin A.
Sample-size calculation reported, 1000 pregnancies per group allowed for a minimum detectable difference of 75 g in birthweight, and >= 34% reduction in fetal loss and >= 45% reduction in infant mortality, with 80% power.
Intention-to-treat analyses performed and the relative risks and confidence intervals were adjusted to account for any cluster design effect.
Compliance: median compliance during pregnancy was 88%.
Location: Salarhi, Nepal.
Timeframe: December 1998 to April 2001.
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Unclear riskUnlcear, 'cluster randomisation using blocks of 5 within each community'.
Allocation concealment?Unclear riskUnlcear, randomisation occurred by 'drawing numbered identical chits from a hat'.
Blinding?
All outcomes
Low riskWomen, field workers and researchers were blinded until after the analysis.
Incomplete outcome data addressed?
All outcomes
Low risk534 (10.7%) women or infants were excluded and 343 (6.7%) infants were lost to follow-up, intention to treat analysis performed.
Free of selective reporting?Unclear riskInformation about women's diet was collected but not reported.
Free of other bias?Unclear riskUnclear due to limited information about the cluster design including allocation concealment.

Correia 1982

Methods

Randomisation and allocation concealment: unclear, "randomised" stated in text but no details given.

Blinding of outcome assessment: "double blind clinical test" stated in the text.

Documentation of exclusion: 16 women (35%) excluded.

Use of placebo control: placebo control.

Participants45 women were initially recruited into the study; however, results are presented for 29 women (folic acid group n = 16, placebo group n = 13). Women were excluded if they had any "pathological data" or if there was "evidence of neglect". No other details given.
InterventionsWomen were randomised to either daily ingestion of 10 mg folic acid or placebo. Women were asked to take the tablets from between 12 and 16 weeks until the end of pregnancy.
Outcomes
  1. Fetal weight (birthweight).

  2. Placental weight.

NotesWomen risk of spontaneous and recurrent miscarriage is unclear. Women's nutritional status is also unclear.
No sample-size calculation reported.
Compliance: unclear, no details given.
Country: Portugal.
Timeframe: unknown.
Published in Portuguese.
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Unclear riskNo methodological details given.
Allocation concealment?Unclear riskNo methodological details given.
Blinding?
All outcomes
Unclear riskUnclear, "double blind clinical test" stated in the text.
Incomplete outcome data addressed?
All outcomes
High risk16 women (35%) excluded.
Free of selective reporting?High riskWomen were excluded if they had any "pathological data" or if there was "evidence of neglect" but not details given.
Free of other bias?Unclear riskLimited methodological details given.

Czeizel 1994

Methods

Randomisation and allocation concealment: unclear, "women agreed to their allocation on the basis of a random table".

Blinding of outcome assessment: unclear, women were aware of the "blind use of one of two kinds of tablets", but no other details given.

Documentation of exclusion: 49 women (1%) were lost to follow-up and excluded.

Use of placebo control: "trace element control" given.

Participants7765 women were recruited into the study. Women participating in the HOFPP who volunteered to take part, were not currently pregnant, and who conceived within 12 months of ceasing contraception. In the first two years of the HOFPP, women were also required to be aged < 35 years, and not to have had a previous pregnancy except a prior induced abortion. 7905 women were approached, of which 140 refused participation, 7765 were randomised and 5502 women had a confirmed pregnancy and were allocated to either multivitamins (n = 2819) or control (n = 2683). 49 women of the 5502 confirmed pregnancies were lost to follow-up.
InterventionsWomen were provided with multivitamin or trace element 'control' from at least 28 days before conception continuing until at least the second missed menstrual period.
The multivitamin with folic acid contained 6000 IU vitamin A, 1.6 mg vitamin B1, 1.8 mg vitamin B2, 2.6 mg vitamin B6, 4.0 mcg vitamin B12, 100 mg vitamin C, 500 IU vitamin D, 15 mg vitamin E, 19 mg nicotinamide, 10 mg calcium pantothenate, 0.2 mg biotin, 0.8 mg folic acid, 125 mg calcium, 125 mg phosphorus, 100 mg magnesium, 60 mg iron, 1 mg copper, 1 mg manganese, 7.5 mg zinc.
The trace element control contained 7.5 mg vitamin C, 1 mg copper, 1 mg manganese and 7.5 mg zinc.
Outcomes
  1. Neural tube defects and other birth defects.

  2. Miscarriage.

  3. Ectopic pregnancy.

  4. Termination of pregnancy.

  5. Live births.

  6. Stillbirths.

  7. Multiple gestation.

  8. Subgroup data is available on menstrual cycle, first trimester symptoms and sexual activity.

NotesWomen's risk of spontaneous and recurrent miscarriage is unclear.
Information on their dietary status is unknown.
No sample-size calculation reported.
Partial intention-to-treat analyses performed.
Compliance: compliance was assessed by questioning, checking the tick-off on the basal temperature chart and counting of unused tablets. 70% of women in the multivitamin group and 71% in the control group took the full course of the supplements, with an additional 20% and 21% in the multivitamin and control groups respectively receiving a partial course of supplementation.
Location: Hungary.
Time frame: 1 February 1984 to 30 April 1992.
The denominators used for this trial are the number of women randomised and with a confirmed pregnancy (i.e. 2819 for the multivitamin group and 2683 for the control group).
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Unclear riskMethodological details unclear.
Allocation concealment?Unclear riskMethodological details unclear, 'women agreed to their allocation on the basis of a random table'.
Blinding?
All outcomes
Unclear riskWomen were aware of the 'blind use of one of two kinds of tablets', but no other details given.
Incomplete outcome data addressed?
All outcomes
Low risk49 women (1%) excluded, partial intention to treat analyses performed.
Free of selective reporting?Unclear riskDenominators vary with serial publications.
Free of other bias?Unclear riskLimited methodological details provided.

Fawzi 1998

Methods

Randomisation and allocation concealment: block randomisation using blocks of 20, eligible women were "assigned the next numbered bottle of regimen". The study used a 2 by 2 factorial design and women were randomised to 1 of 4 groups. Tablets were indistinguishable and packaged in identically coded bottles.

Blinding of outcome assessment: women and study investigators were unaware of the treatment allocation, no information given about blinding of outcome assessors.

Documentation of exclusion: 64 women (6%) were lost to follow-up and excluded.

Use of placebo control: placebo given.

Participants1085 women were recruited into the study. Pregnant women between 12 and 27 weeks' gestation who were HIV-1 infected, living in Dar es Salaam and intended to stay there for at least one year were eligible for the study. Women not HIV-1 positive or moving out of Dar es Salaam were excluded. 13,879 pregnant women consented to be HIV-1 tested, of which 1806 were positive, and 1085 were randomised. Of these, 3 women were not pregnant and 7 women died before delivery and were excluded from the trial. Of the remaining 1075 women, 54 women (5%) were lost to follow-up by the time of delivery, leaving birth outcomes reported for 1021 women. Women were randomised to 1 of 4 groups: vitamin A (n = 269), multivitamins excluding vitamin A (n = 269); multivitamins including vitamin A (n = 270) or placebo (n = 267).
Interventions

Women were randomised to 1 of 4 groups:

  1. vitamin A (30 mg beta-carotene plus 5000 IU preformed vitamin A);

  2. multivitamins excluding vitamin A (20 mg vitamin B1, 20 mg vitamin B2, 25 mg vitamin B6, 100 mg niacin, 50 mcg vitamin B12, 500 mg vitamin C, 30 mg vitamin E, 0.8 mg folic acid);

  3. multivitamins including vitamin A, all formulated in 2 tablets; or

  4. placebo.

All women received 400 mg ferrous sulphate and 5 mg folic acid daily, as well as 500 mg chloroquine phosphate weekly. At delivery, all women taking vitamin A were to receive an additional oral dose of 200,000 IU vitamin A and the others an extra dose of a placebo. Pill counts were conducted at each visit and new tablets were given out at each visit.

Outcomes
  1. Miscarriage, defined as delivery before 28 weeks' gestation.

  2. Stillbirth, defined as delivery of a dead baby at or after 28 weeks' gestation.

  3. Fetal death, defined as either miscarriage or stillbirth.

  4. Low birthweight, defined as birthweight less than 2500 g.

  5. Very low birthweight, defined as birthweight less than 2000 g.

  6. Preterm delivery, defined as delivery before 37 weeks.

  7. Severe preterm birth, defined as delivery before 34 weeks.

  8. Small-for-gestational age, defined as birthweight less than the 10th percentile for gestational age.

NotesWomen's risk of spontaneous and recurrent miscarriage was unclear, although may be increased due to their HIV-1 positive status.
Women's nutritional status is also unclear.
Figures change with serial publications, particularly for secondary outcomes, and results are not reported separately for the individual 4 groups. Results are reported as: any multivitamins, multivitamin, any vitamin A or no vitamin A.
Sample-size calculation performed allowing for 20% loss to follow up.
Intention-to-treat analyses performed.
Compliance: compliance assessed by the percentage of prescribed tablets absent from the returned bottles, and in plasma vitamin A concentrations in a subset of 100 women. Median compliance assessed using pill counts was 90% by the time of delivery.
Location: Tanzania.
Timeframe: April 1995 to July 1997.
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Unclear riskBlock randomisation using blocks of 20.
Allocation concealment?Unclear riskWomen assigned the 'next numbered bottle of regimen'.
Blinding?
All outcomes
Low riskWomen and investigators were blinded.
Incomplete outcome data addressed?
All outcomes
High risk64 women (6%) were lost to follow up and excluded, intention to treat analyses performed.
Free of selective reporting?Unclear riskFigures change with serial publications, particularly for secondary outcomes, and results are not reported separately for the individual 4 groups.
Free of other bias?Unclear riskLimited methodological details provided.

Fawzi 2007

Methods

Randomisation: unclear about sequence generation.

Allocation concealment: States a list was prepared according to the randomisation sequence in blocks of 20, tablets were bottled in identical coded bottles, eligible women were given the next numbered bottle.

Blinding of outcome assessment: Women and research assistants who assessed the study outcomes were unaware of the intervention groups.

Documentation of exclusion: 49 women lost to follow-up (multivitamin group: 23, placebo group: 26), no post-randomisation exclusions.

Use of placebo control: placebo given.

Participants8428 women were randomised in the study. Pregnant women between 12 and 27 weeks who had a negative test for HIV infection and planned to stay in the city until delivery and for 1 year thereafter recruited through antenatal clinics in Dar es Salaam. 8468 women were enrolled, however 40 women were then found to be ineligible. 8428 women were randomly assigned to receive either a multivitamin (n = 4214) or placebo (n = 4214) from the time of enrolment until 6 weeks after delivery. 6 women died before delivery and 43 were lost to follow up by the time of delivery.
Interventions

The supplements included 20 mg of vitamin B1, 20 mg of vitamin B2, 25 mg of vitamin B6, 100 mg of niacin, 50 mcg of vitamin B12, 500 mg of vitamin C, 30 mg of vitamin E, and 0.8 mg of folic acid.

The active tablets and placebo were similar in shape, size, and colour.

All women, irrespective of the assigned study regimen, were given daily doses of iron (60 mg of elemental iron) and folic acid (0.25 mg). They were also given malaria prophylaxis in the form of sulfadoxine-pyrimethamine tablets at 20 weeks and 30 weeks of gestation.

Outcomes
  1. Low birthweight (< 2500 g).

  2. Preterm delivery (before 37 weeks' gestation).

  3. Fetal death.

  4. Birthweight below 2000 g.

  5. Extremely preterm delivery (before 34 weeks).

  6. Small-for-gestational age (birthweight below the 10th percentile for gestational age).

  7. Fetal death and death in the first 6 weeks of life.

Notes

Women's risk of spontaneous and recurrent miscarriage was unclear.

Women's nutritional status is also unclear.

Intention to treat analyses performed.

Compliance: Average compliance was 88%, no difference in compliances between the two groups.

Location: Tanzania.

Timeframe: August 2001 and July 2004.

Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Unclear riskGeneration of sequence not reported, except that there were blocks of 20 in the sequence.
Allocation concealment?Low riskIdentical coded bottles prepared according to the randomisation list, eligible women were assigned the next numbered bottle.
Blinding?
All outcomes
Low riskWomen and outcome assessors were blinded to allocation.
Incomplete outcome data addressed?
All outcomes
Low risk49 (1%) women lost to follow-up, balanced across groups, analyses by intention to treat.
Free of selective reporting?Low riskAll pre-specified outcomes appear to be reported.
Free of other bias?Low riskThe study appears to be free of other sources of bias.

Fleming 1968

Methods

Randomisation and allocation concealment: quasi-randomised, alternate women were allocated to receive folic acid or placebo according to the order in which they attended antenatal clinic. No other methodological details were given.

Blinding of outcome assessment: women and investigators were blinded to the treatment allocation, until after the completion of the trial.

Documentation of exclusion: 21 women (28%) excluded from the analysis.

Use of placebo control: control tablet containing iron given.

Participants75 women were recruited into the trial. Women were eligible if they were primigravida, less than 26 weeks' pregnant (range of gestation 10 to 26 weeks'), with haematocrit value (PCV) 27 per cent or more, and who had not received treatment so far as was known. Women with Haemoglobin (Hb) SC, Hb.SS, Hb.CC were excluded. Alternate patients were allocated to group A (placebo) or B (folic acid). 75 women were included (40 in group A and 35 in group B) initially; however, only 26 in group A and 28 in group B completed the trial. 16 women (10 in group A and 8 in group B) defaulted from the trial, 3 (2 in group A and 1 in group B) were anaemic on the second visit warranting folic acid treatment, 1 in group A self medicated with folic acid and 1 in group A 'aborted'.
InterventionsAll women received antimalarials and iron supplements as per the standard antenatal care at the hospital.
Women in group B received 5 mg folic acid tablets on each attendance, which was fortnightly initially and weekly in the last trimester.
Group A received "one tablet of lactose base and colouring matter in the same manner."
Outcomes
  1. PCV and reticulocyte index.

  2. Serum folic acid concentration and 'megaloblastic score'.

  3. Malarial infection.

  4. Maternal morbidity (pyelonephritis, pre-eclamptic toxaemia, septicaemia, puerperal psychosis).

  5. Prematurity.

  6. Birthweight (mean birthweight but no standard deviation).

  7. Fetal mortality.

NotesResults not reported as intention to treat; however, where possible, the review authors included data in the review as intention to treat.
Unclear of women's risk of spontaneous and recurrent miscarriage.
16 women in the trial showed evidence of folic acid deficiency at trial entry.
Sample-size calculation: none reported.
No intention-to-treat analyses performed.
Compliance: no compliance information reported specifically; however, women were "seen to swallow" the tablets at their fortnightly and weekly visits.
Location: Nigeria.
Time frame: unclear.
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?High riskQuasi-randomised, alternate allocation.
Allocation concealment?High riskQuasi-randomised, alternate allocation.
Blinding?
All outcomes
Low riskWomen and investigators blinded.
Incomplete outcome data addressed?
All outcomes
High risk21 women (28%) excluded from the analysis.
Free of selective reporting?Unclear riskResults not reported as intention to treat; however, where possible, the review authors included data in the review as intention to treat.
Free of other bias?Unclear riskLimited methodological details provided.

Fleming 1986

Methods

Randomisation and allocation concealment: women were "randomly allocated to one of five groups using a random number table", no other details given.

Blinding of outcome assessment: Women and investigators were blinded to the treatment allocation, until after the completion of the trial.

Documentation of exclusion: 18 women (9%) were excluded due to anaemia at enrolment, 'defaulting', or being 'mentally subnormal', these women were replaced by other women chosen by an investigator. A further 42 women were excluded before delivery and another 30 failed to attend the postnatal clinic, birth outcomes were available for 160 women (80%).

Use of placebo control: no placebo control.

Participants

228 women met the eligibility criteria; however 200 pregnant women were recruited into the study. Women were allocated to one of five groups; 40 women were allocated to each group.

Eligible women included:

  1. Hausa women living in Zaria and planning to deliver in Zaria;

  2. pregnant for the first time;

  3. at less than 24 weeks' gestation, as estimated by the height of the fundus uteri;

  4. the wives of unskilled or semiskilled men.

Women were excluded if they had already taken any antimalarial treatment or haematinics during the pregnancy, or had the following complications: hydatiform mole, haemoglobin SC disease, overt anaemia or proteinuria.

The mean gestational age of women at enrolment was 18.5 weeks.

Interventions

Women were allocated to 1 of 5 groups:

  • Group 1: No active treatment (control);

  • Group 2: Antimalarials only (600mg chloroquine/day + 100 mg proguanil/day);

  • Group 3: Iron + antimalarials (60 mg iron/day + 600mg chloroquine/day + 100 mg proguanil/day);

  • Group 4: Folic acid + antimalarials (1 mg folic acid/day + 600mg chloroquine/day + 100 mg proguanil/day);

  • Group 5: Iron + folic acid + antimalarials (1 mg folic acid/day + 60 mg iron/day + 600mg chloroquine/day + 100 mg proguanil/day).

Outcomes

Maternal outcomes

  1. Anaemia (severe and mild/moderate) before 28 weeks', between 28-36 weeks', and after 36 weeks' gestation.

  2. Gestation age.

  3. Mode of delivery.

  4. Complications of pregnancy (abortion, hypertension, pre-eclampsia or eclampsia, hydramnios, abdominal pain).

Infant outcomes

  1. Fetal distress.

  2. Birthweight.

  3. Apgar score at two minutes.

  4. Fetal complications.

Laboratory outcomes

  1. Hb concentration, red cell indices and WBC at first attendance, 28 weeks, 36 weeks, at delivery (form mother and infant) and six weeks postpartum.

Not all outcomes were reported for each individual treatment group. Miscarriage was reported for the combined groups 4 and 5, therefore for the purpose of this review the groups 4 and 5 are combined (folic acid + iron) and compared with group 2 and group 3 (iron + antimalarials). The authors reported that 8 women had hypertension without other signs, 21 women had preeclampsia and 6 developed eclampsia, with no association between these outcomes and treatment group. No other details were provided, including the breakdown of these outcomes by treatment group.

Notes

Women's risk of spontaneous and recurrent miscarriage was unclear.

Women were at high risk of anaemia. Information about other nutritional indices was not provided.

Intention to treat analyses not performed, however, where possible, the review authors included data in the review as intention to treat.

Compliance: 72 women (36%) were classed as defaulters.

Location: Nigeria.

Timeframe: Unclear.

Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Unclear riskA random number table was used but no details provided of how it was generated.
Allocation concealment?Unclear riskNo details provided about the allocation.
Blinding?
All outcomes
Low riskNeither the researchers nor the patients were aware of the treatment allocation until after the completion of the study.
Incomplete outcome data addressed?
All outcomes
High risk228 women met the entry criteria, but only 200 were included in the trial. 18 women were excluded and replaced by other women.
Free of selective reporting?High riskNot all outcomes are reported by treatment group. In serial publications up to 70% of the data was excluded.
Free of other bias?Unclear riskLimited methodological details provided.

Hemmi 2003

Methods

Randomisation and allocation concealment: unclear, "patients were randomly assigned to the control group or the study group". No other methodological details given.

Blinding of outcome assessment: unclear, no details given.

Documentation of exclusion: 28 women (19%) in the control group were excluded, no details given for the exclusion.

Use of placebo control: no placebo control.

Participants150 women were recruited into the study. Women with a luteal phase defect, as described by a peak serum P level < 120 mg/mL in the mid-luteal phase measured at 3 time points, were eligible and invited to participate. Luteal phase defects were ascertained in two consecutive menstrual cycles, and the third cycle was the intervention cycle. Women receiving IVF-ET treatment were excluded. 313 women were considered for enrolment in the study, 150 (48%) were randomised. 28 women were withdrawn from the control group, leaving 122 women in the study, who were allocated to vitamin C (n = 76) or control (n = 46). 5 women in the control group and 19 women in the vitamin C group became pregnant during the study period.
InterventionsWomen in the intervention group took 750 mg vitamin C per day from the first day of the third menstrual cycle until a urinary pregnancy test was positive. Pregnancy rate was checked up until 6 months after the study cycle was started. Women in the control group received no supplementation and no treatment was given in the third cycle.
Outcomes
  1. Serum P concentrations.

  2. Serum E2 (oestrogen) concentrations.

  3. Pregnancy rate.

  4. Miscarriage.

NotesWomen's risk of spontaneous or recurrent miscarriage was unclear according to criteria specified in the review.
Their dietary intake of vitamin C is unknown.
No sample-size calculation was reported.
Analyses were not based on intention to treat.
Compliance: no details of any compliance assessments were given.
Country: Japan.
Time frame: January 1997 to December 2000.
The denominators used for this trials are the number of women randomised and with a confirmed pregnancy (i.e. 19 for the vitamin group and 5 for the control group).
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Unclear riskMethodological details unclear.
Allocation concealment?Unclear riskMethodological details unclear.
Blinding?
All outcomes
Unclear riskMethodological details unclear.
Incomplete outcome data addressed?
All outcomes
High risk28 women (19%) in the control group excluded.
Free of selective reporting?Unclear riskNo details of exclusion of women in the control group given.
Free of other bias?High riskNo placebo control.

ICMR 2000

Methods

Randomisation and allocation concealment: unclear, "containers of vitamin or placebo capsules were given a random number" and "the key to random numbers was kept at the ICMR Headquarters". No other methodological details were given.

Blinding of outcome assessment: "double blind" mentioned in the text, but no details given.

Documentation of exclusion: 187 women (40%) were excluded from the analysis.

Use of placebo control: placebo control.

Participants466 women were recruited into the study. Women who had previously given birth to a child with an open NTD, and planned to have another child were eligible and invited to participate. This was regardless of their parity, number of previous births with an NTD, age, consanguinity, and socio-economic status. Women who had previously given birth to a child with closed spina bifida, or with a history of diabetes or abnormal fasting and post-prandial blood sugar, history of epilepsy, congenital anomalies indicative of a genetic syndrome in the previous NTD, history of vitamin intake in the 3 months prior to enrolment, and pregnancy were excluded. 466 women were enrolled and randomised to either vitamin (n = 231) or placebo (n = 235), of these women, 90 were lost to follow-up immediately and 71 did not conceive until the final follow-up. Of the remaining 305 women who were known to become pregnant (vitamin n = 152, placebo n = 153), pregnancy outcomes were unknown for 26 women. In the paper, 279 of the initial 466 women were included in the analysis; however, in this review results are presented for main outcomes on an intention-to-treat basis (i.e. n = 466).
InterventionsThe folic acid containing multivitamin included 120 mg ferrous sulphate, 240 mg calcium phosphate, 4000 IU vitamin A, 400 IU vitamin D, 2.5 mg vitamin B1, 2.5 mg vitamin B2, 2 mg vitamin B6, 15 mg nicotinamide, 40 mg vitamin C, 4 mg folic acid, 10 mg zinc.
The placebo tablets contained the following trace elements: 120 mg ferrous sulphate and 240 mg calcium phosphate. Both capsules were identical in appearance and women were provided with the tablets from at least 28 days before conception and continuing until at least the second missed menstrual period.
Outcomes
  1. Recurrence of neural tube defects.

  2. Live births.

  3. Stillbirths.

  4. Spontaneous and induced abortion.

  5. Multiple birth.

NotesThe risk profile of women in the trial for spontaneous and recurrent miscarriage is unclear, as is the dietary intake of participants.
Sample-size calculation performed, assuming a 20 per cent drop out rate. The trial was terminated after publication of the MRC trial in 1991.
Compliance: compliance was assessed at 3 monthly visits, by checking a diary card maintained by the woman and the number of capsules returned. If the total number of missed days in 3 months did not exceed 10 days, and the total number of missed days at a stretch did not exceed three, compliance was taken as satisfactory. Women not meeting the above criteria were excluded if they became pregnant in that particular quarter. No compliance data are specifically reported.
Analyses not based on intention to treat.
Country: India.
Time frame: 1988 to 1991.
The denominators used for this trial are based on the number of women randomised (i.e. 231 for the vitamin group and 235 for the placebo group). There was not enough information to accurately confirm the number of women that did or did not become pregnant due to the large number of losses to follow up.
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Unclear riskContainers 'given a random number'.
Allocation concealment?Unclear risk'Key to random numbers were kept at the ICMR headquarters' but no other details given.
Blinding?
All outcomes
Unclear riskDouble blind mentioned in the text but no details given.
Incomplete outcome data addressed?
All outcomes
High risk187 (40%) women excluded.
Free of selective reporting?Unclear riskDifficult to assess given the high losses to follow-up.
Free of other bias?Unclear riskLimited methodological details provided.

Katz 2000

Methods

Randomisation and allocation concealment: cluster randomised. 270 centres in the Salarhi district, Nepal, were involved which included 30 subdistricts each with 9 wards. Each ward was assigned to 1 of 3 treatment groups. "Wards were assigned by a random draw of numbered chits, blocked on subdistrict".

Blinding of outcome assessment: women and study investigators were not aware of the treatment codes. Maternal mortality was assessed by study investigators blinded to treatment allocation, no details were given for other outcomes.

Documentation of exclusions: 157 (1%) women were lost to follow-up and excluded.

Use of placebo: placebo control.

Participants15,832 women were recruited into the study. All married women of child bearing age in the Salarhi district, Nepal, were eligible and invited to participate in the study. Women migrating into the study area, or women that were never pregnant or refused participation, or women who migrated before being pregnant, were excluded from the analysis. Eligible women were identified from census data and marriage registers. 44,646 women were recruited, of which 1136 (2.5%) were excluded as they either emigrated before becoming pregnant, died or refused consent. During the study period 15,832 women identified themselves as being pregnant, and 157 women were lost to follow-up in the postpartum period. Results are reported for 17,373 pregnancies, allocated to the following groups: vitamin A (n = 6070), beta-carotene (n = 5650) or placebo (n = 5653). Denominators for the treatment groups vary for the measures of early infant mortality, due to losses to follow-up after birth.
Interventions

The three treatment groups consisted of a weekly single oral supplement of either:

  1. 23,300 IU preformed vitamin A as retinyl palmitate;

  2. 42 mg of all trans beta-carotene;

  3. placebo.

All capsules contained mg dl-alpha-tocopherol as an antioxidant. Women took the tablets prior to conception, during pregnancy and postpartum, for a total of 3.5 years.

Outcomes1. Fetal loss, defined as any reported miscarriage, stillbirth or maternal death during pregnancy. The outcomes were based on self reports, and women who reported to be pregnant for >= 6 weeks but then no longer reported being pregnant were considered to have had a miscarriage.
Serial publications also reported neonatal death.
NotesWomen's risk profile for spontaneous or recurrent miscarriage was unclear, as was their dietary intake of vitamin A.
Compliance: women were distributed the capsules in their home on a weekly basis, receipt of capsules was noted only if the distributor observed the woman swallowing the capsule. Over half of the women who became pregnant during the study received over 80% of their intended supplements, and 75% of pregnant women received at least half of their eligible doses.
There were serial publications of this study causing the study numerators and denominators to vary between published versions, and multiple pregnancy figures reported did not include higher order pregnancies.
Sample-size calculation performed.
Partial intention-to-treat analyses, and the relative risks and confidence intervals were adjusted to account for any cluster design effect.
Country: Nepal.
Timeframe: April 1994 to September 1997.
The denominators used for this trial are the number of women randomised who identified themselves as pregnant (i.e. 6070 for the vitamin A group, 5650 for the beta-carotene group and 5653 for the placebo group).
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Unclear riskCluster randomised, unclear how sequence was generated.
Allocation concealment?Unclear riskEach ward was assigned to the treatment groups based on 'a random draw of numbered chits, blocked on subdistrict'.
Blinding?
All outcomes
Low riskWomen and investigators blinded to treatment allocation.
Incomplete outcome data addressed?
All outcomes
Unclear risk157 women (1%) were lost to follow-up and excluded, partial intention to treat analysis performed.
Free of selective reporting?High riskDenominators vary in serial publications of this trial.
Free of other bias?High riskSome women were pregnant more than once during the study period, however the denominators reported are the total number of pregnancies during the study period, not the total number of women randomised, which incorrectly assumes that each data point included is independent from the next.

Kirke 1992

Methods

Randomisation and allocation concealment: block randomisation, stratified by hospital, using "consecutively numbered, opaque, sealed envelopes".

Blinding of outcome assessment: women and study investigators were initially blinded to the treatment allocation, however the tablet preparations were changed after 55 women were randomised and after this only participants were blinded.

Documentation of exclusion: 3 women (1%) were lost to follow-up and excluded.

Use of placebo control: 3 treatment regimens were assessed, no placebo control.

Participants354 women were recruited into the study. Women with a previous neural tube defect defined as anencephalus, iniencephalus, encephalocoele, and spina bifida aperta, who were not pregnant when contacted but were planning a future pregnancy, were eligible and invited to participate. Women were identified from case registers at the participating hospitals. Women with conditions likely to result in impaired absorption from the gastrointestinal tract were excluded.
435 women were approached, of which 354 (84%) consented and were randomised to either F (n = 115 ), MV (n = 119) or MF (n = 120). 16 women did not become pregnant, and 75 women withdrew; however, their pregnancy outcome status was known, and 18 of these women subsequently became pregnant after withdrawing. 3 women were lost to follow-up. 281 women (93 in the F group, 93 in the MF group and 95 in the MV group) became pregnant in the study period and their pregnancy outcome was known.
Interventions

Indistinguishable trial tablets were initially made by Beecham and Glaxo, however Beecham withdrew their support after 55 women had been randomised. After this time a commercially available pregnavite Forte F was used (MF tablet) and Antigen Pharmaceuticals produced a white multivitamin tablet without folic acid. This was associated with a loss of blinding. Women were randomised to one of three treatments:

  1. folic acid alone (F);

  2. multivitamin with folic acid (MF);

  3. multivitamin with no folic acid (MV).

The F and MF resulted in a daily dose of 0.3 mg folic acid. The MF and MV resulted in a daily dose of 4000 IU vitamin A, 400 IU calciferol, 1.5 mg thiamine hydrochloride, 1.5 mg riboflavine, 1 mg pyridoxine hydrochloride, 15 mg nicotinamide, 40 mg ascorbic acid, 480 mg calcium phosphate, and 252 mg ferrous sulphate. Women took the tablets for at least 2 months prior to conception and until the date of the 3rd missed period.

Outcomes
  1. Recurrence risk of neural tube defects.

  2. Spontaneous abortion.

  3. Ectopic pregnancy.

  4. Livebirth.

  5. Stillbirth.

  6. Congenital malformations excluding neural tube defects.

NotesThe trial was stopped after there were poor recruitment rates and birth rates. A sample-size calculation required 462 women to show a reduction in neural tube defects from 5% to 1%. Data from 106 women who were already pregnant at time of recruitment are also included.
The risk profile of women in the trial for spontaneous and recurrent miscarriage is unclear, as is their dietary intake.
Compliance: compliance was assessed on tablet counts and blood tests; however, the results are not presented.
Intention-to-treat analyses were performed.
Country: Republic of Ireland.
Timeframe: December 1981 to January 1988.
The denominators used for this trial are the number of women randomised who became pregnant in the study period and their pregnancy outcome was known (i.e. 93 in the F group, 93 in the MF group and 95 in the MV group).
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Low riskBlock randomisation stratified by hospital site.
Allocation concealment?Low riskConsequtively numbered, opaque sealed envelopes used.
Blinding?
All outcomes
Low riskOnly participants were blinded.
Incomplete outcome data addressed?
All outcomes
Low risk3 women (1%) lost to follow-up and excluded. Intention to treat analyses performed.
Free of selective reporting?Unclear riskCompliance data not reported.
Free of other bias?High riskThe trial was stopped after there were poor recruitment rates and birth rates.

Kumwenda 2002

Methods

Randomised controlled trial of vitamin A, iron and folic acid supplementation versus iron and folic acid only, during pregnancy, to improve infant outcomes born to women infected with HIV in Malawi.

Randomisation and allocation concealment: "treatment assignment was determined by use of a computer's random-number generator" and "mothers were assigned an original study identification number at enrolment and were given the next sequentially numbered opaque bottle with supplements". "Treatment assignment was concealed by pre packing study supplements in sequentially numbered series assigned to study identification numbers."

Blinding of outcome assessment: unclear, not specifically stated, but participants were blind to their treatment allocation.

Documentation of exclusion: 63 (9%) women were lost to follow-up and 14 (2%) pairs of twins were excluded.

Use of placebo control: control tablets containing iron and folic acid were given.

ParticipantsPregnant women between 18 and 29 weeks' gestation and infected with HIV. The average gestation of participants was 23 weeks. 693 women were enrolled and allocated to either vitamin A (n = 340) or control (n = 357), of which pregnancy outcomes were known for 623 women. 63 women were lost to follow-up and 14 sets of twins were excluded due to their higher risk of low birthweight and infant mortality.
InterventionsAll women received orally administered daily doses of 30 mg iron and 400 mcg folic acid during the study. Women in the intervention group received 10,000 IU vitamin A (3 mg retinol equivalent) orally, in addition to the iron and folic acid supplements. Women were asked to take the tablets from enrolments until delivery. Tablet counts were conducted every 4 weeks. All women received 30 mg retinol equivalents at 6 weeks postpartum, according to standard postpartum care in Malawi.
Outcomes
  1. Infant haemoglobin level at 6 weeks and 12 months of age.

  2. Percentage of infants with anaemia at 6 weeks of age and at 12 months, defined as a haemoglobin level of < 110 g/L.

  3. Birthweight.

  4. Percentage of infants < 2500 g at birth.

  5. Weight and length at 6 weeks, 14 weeks and 6 months of age.

  6. Transmission of HIV to the infant, infant mortality at < 6 weeks of age, at 12 months and at 24 months.

  7. Stillbirth and spontaneous abortion (undefined).

NotesWomen's risk of spontaneous and recurrent miscarriage is unclear, although may be increased due to their HIV status.
50% of women in the vitamin A group and 51% of women in the control group had deficient levels of vitamin A (defined as plasma vitamin A < 0.70 umol/L) at trial entry.
Sample-size calculation performed.
No intention-to-treat analyses were performed.
Compliance: more than 95% of women in both groups took > 90% of study supplements, as ascertained by tablet counts.
Location: Malawi.
Timeframe: November 1995 toDecember 1996.
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Low riskComputer generated random number list.
Allocation concealment?Low riskSequentially number opaque bottles used.
Blinding?
All outcomes
Unclear riskNot specifically stated but women were blinded.
Incomplete outcome data addressed?
All outcomes
Unclear risk63 women (9%) lost to follow-up and 14 pairs of twins (2%) excluded. No intention to treat analyses performed.
Free of selective reporting?Low riskAll pre-specified outcomes appear to be reported.
Free of other bias?Unclear riskInsufficient information to assess whether an important risk of other bias exists.

MRC 1991

Methods

Randomisation and allocation concealment: third party randomisation, "randomisation was carried out through the Clinical Trials Service Unit in Oxford". Randomisation was stratified by centre.

Blinding of outcome assessment: women, caregivers and study investigators were blinded to the treatment allocation.

Documentation of exclusion: 164 women (9%) excluded.

Use of placebo control: placebo control.

Participants1817 women were recruited into the study. Women who had a previous pregnancy affected by a neural tube defect, and were planning another pregnancy and not already taking supplements were eligible for the study. Women whose affected child had Meckel's syndrome and those women with epilepsy were excluded. 1817 women were randomised to either F (n = 449), MV (n = 453), MF (n = 461) or P (n = 454), of which, 1195 were informative pregnancies that is, where the outcome of NTD or not was definitely known (F n = 298, MV n = 302, MF n = 295, P n = 300). Results for pregnancy loss are reported for both informative and not informative pregnancies. 164 women were excluded as they may have been pregnant at the time of randomisation.
Interventions

Women were randomised into 1 of 4 groups:

  1. 4 mg, 240 mg di-calcium phosphate and 120 mg ferrous sulphate (F);

  2. 4000 IU vitamin A, 400 IU calciferol, 1.5 mg thiamine hydrochloride, 1.5 mg riboflavine, 1 mg pyridoxine hydrochloride, 15 mg nicotinamide, 40 mg ascorbic acid, 240 mg di-calcium phosphate and 120 mg ferrous sulphate (MV);

  3. folic acid combined with the multivitamins specified above (MF);

  4. placebo containing 240 mg di-calcium phosphate and 120 mg ferrous sulphate only (P).

Women took the tablets prior to conception and attended the site every 3 months to collect additional supplies and again during the 12th week of pregnancy. No special dietary advice was given to women.

Outcomes
  1. Neural tube defect and other birth defects.

  2. Spontaneous abortions.

  3. Ectopic pregnancy.

  4. Termination or pregnancy.

  5. Livebirth.

  6. Stillbirth.

  7. Multiple pregnancy.

  8. Subsequent publications report on blood folic acid and zinc concentrations.

NotesThe trial was stopped early after there were 1195 informative pregnancies, according to prespecified stopping rules. The aim of the study was to obtain information on at least 2000 informative pregnancies unless a sufficiently clear result emerged sooner.
Women's risk profile for spontaneous and recurrent miscarriage was unclear, as was their nutritional status.
Compliance: compliance based on self reports, and data were available for women with an informative pregnancy only, where 79 (6%) women reported they stopped taking their capsules before their last scheduled visit.
Intention-to-treat analyses are reported in this review including not informative pregnancies (i.e. n = 1817).
Location: multi-national study coordinated from the United Kingdom.
Timeframe: July 1983 to April 1991.
The denominators used for this trial are the number of women randomised i.e. (449 for the F group, 453 for the MV group, 461 for the MF and 454 for the P group). There was no information provided about any women randomised that did not become pregnant in the study period.
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Low riskThird party randomisation, "randomisation was carried out through the Clinical Trials Service Unit in Oxford".
Allocation concealment?Low riskThird party randomisation, "randomisation was carried out through the Clinical Trials Service Unit in Oxford".
Blinding?
All outcomes
Low riskWomen, caregivers and investigators blinded to treatment allocation.
Incomplete outcome data addressed?
All outcomes
Low risk164 women (9%) excluded, intention-to-treat analyses performed.
Free of selective reporting?Unclear riskNo information provided about any women randomised that did not become pregnant in the study period.
Free of other bias?High riskThe trial was stopped early after there were 1195 informative pregnancies, according to prespecified stopping rules.

Osrin 2005

Methods

Randomisation and allocation concealment: One of the authors 'randomly allocated 1200 participant numbers by computer into two groups in permuted blocks of 50'. Every identification number was allocated a supplement container, which was then packed by a team member not otherwise involved in the trial. After enrolment, another author allocated participants sequential identification numbers with the corresponding supplement containers.

Blinding of outcome assessment: double blind stated but no other details given.

Documentation of exclusion: 61 women (5%) withdrew or were lost to follow-up, however data on miscarriage were reported for those who withdrew due to miscarriage.

Use of placebo control: control of iron and folic acid supplements given which looked identical to the intervention supplements.

Participants

1200 women were recruited into the study. Women were eligible if they were: less than 20 completed weeks, had a singleton pregnancy, no notable fetal abnormality, no existing maternal illness of a severity that could compromise the outcome of pregnancy, and lived in an area of Dhanusha or the adjoining district of Mahottari accessible for home visits.

Maternal illnesses that led to exclusion were: recently treated recurrent cysticercosis, need for chlorpromazine or anticoagulant drugs with changing doses, and symptomatic mitral stenosis or multivalvular heart disease. Fetal exclusions were: twin pregnancies, anencephaly, occipital meningocele, encephalocele, duodenal atresia and a grossly dilated pelvicalyceal system.

Interventions

Intervention group: vitamin A 800 µg, vitamin E 10 mg, vitamin D 5 µg, vitamin B1 1.4 mg, vitamin B2 1.4 mg, niacin 18 mg, vitamin B6 1.9 mg, vitamin B12 2.6 µg, folic acid 400µg, vitamin C 70 mg, iron 30 mg, zinc 15 mg, copper 2 mg, selenium 65 µg, and iodine 150 µg.

Control group: iron 60 mg and folic acid 400 µg.

Supplementation began at a minimum of 12 weeks’ gestation and continued until delivery.

Outcomes
  1. Birthweight.

  2. Gestational duration.

  3. Infant length and head circumference.

  4. Miscarriage defined as cessation of confirmed pregnancy before 23 weeks’ gestation.

  5. Stillbirth defined as delivery of an infant showing no signs of life (movement, breathing, or heartbeat) after 23 weeks’ gestation.

  6. Early neonatal death defined as death of a live born infant in the first 7 days after birth.

  7. Late neonatal death as death of a live born infant after 7 but within 28 days.

Notes

Women's risk of spontaneous and recurrent miscarriage was unclear.

Women's nutritional status is also unclear, however, women are presumable at high risk of under-nutrition as the paper states that in Nepal 'deficiencies of several micronutrients have been well described in individual studies and in a national sample'.

Intention to treat analyses performed.

Compliance: Median 'adherence' was 98% in the control group and 97% in the intervention group.

Location: Nepal.

Timeframe: August 2002 to October 2003.

Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Low riskComputer generated in permuted blocks of 50.
Allocation concealment?Unclear riskOne of the authors allocated participants with sequential identification numbers, but unclear if this person was involved in the recruitment of participants.
Blinding?
All outcomes
Unclear riskDouble blind stated in the text but no other details given.
Incomplete outcome data addressed?
All outcomes
Low risk61 women (5%) withdrew or were lost to follow-up, however data on miscarriage were reported for those who withdrew due to miscarriage. Intention-to-treat analyses performed.
Free of selective reporting?Low riskAll pre-specified outcomes appear to be reported.
Free of other bias?Low riskThe study appears to be free of other sources of bias.

People's League 1942

Methods

Randomisation and allocation concealment: "women enrolled at the antenatal clinic were divided into two main groups by placing them alternatively on separate lists".

Blinding of outcome assessment: unclear, no information given on blinding of participants, carers or outcome assessors.

Documentation of exclusion: 622 (11%) women were excluded.

Use of placebo control: no placebo given.

Participants5644 women were recruited into the study. All women attending the antenatal clinics and who were less than or equal to 24 weeks' gestation and who were in 'good health' were eligible for the study. Women who were more than 24 weeks' gestation and women who suffered from any disease or physical abnormality were excluded from the study. After enrolment, women who had twin births and who miscarried at an early stage were also excluded.
5644 women were initially enrolled in the study of which 5022 (89%) remained in the study. Of the 622 (11%) women withdrawn from the trial, 494 were evacuated from the London area (due to World War 2), 39 women had twin births and 89 women miscarried at an early stage. 5022 women remained in the study and were allocated to either multivitamins (n = 2510) or control (n = 2512). Women were further divided into primiparae and multiparae, and various age groups.
InterventionsWomen allocated to the treatment group were given daily vitamin C 100 mg, ferrous iron 0.26 g, calcium 0.26 g, minute quantities of iodine, manganese and copper, adsorbate of vitamin B1 containing all factors of the B complex and halibut liver oil 0.36 g containing vitamin A (52,000 IU per g) and vitamin D (2500 IU per g).
Women allocated to the control group received no placebo.
Outcomes
  1. Toxaemia classified into subgroups based on: hypertension only, albuminuria with or without hypertension, or hypertension with albuminuria (pre-eclampsia).

  2. Maternal sepsis.

  3. Length of gestation (categorised as less than 40 weeks, 40 weeks, and greater than 40 weeks).

  4. Percentage of women breastfeeding.

  5. Stillbirth.

  6. Neonatal mortality (defined as death before 8 days).

  7. Birthweight (pounds) (only reported for primiparae and multiparae separately).

NotesWomen risk status for spontaneous and recurrent miscarriage is unclear.
Dietary intake at trial entry: "vitamin C shortage affected about half the women".
Intention-to-treat analyses: not performed.
Compliance: unclear, no information provided.
Sample-size calculation: unclear. "It was decided that the investigation should include a minimum of 5000 pregnant women". No other details given.
Location: England.
Timeframe: 1938 to 1939.
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?High riskQuasi-randomistion using alternate separate lists.
Allocation concealment?High riskNo allocation concealment.
Blinding?
All outcomes
Unclear riskNo information about blinding provided.
Incomplete outcome data addressed?
All outcomes
Unclear risk622 women (11%) excluded, intention to treat analyses not performed.
Free of selective reporting?Unclear riskLimited methodological details provided.
Free of other bias?Unclear riskLimited methodological details provided.

Roberfroid 2008

Methods

Randomisation and allocation concealment: the randomisation scheme was generated by a computer program in permuted blocks of 4. Randomisation numbers were sealed in opaque envelopes. At each inclusion, the consulting physician opened the next sealed envelope and transmitted the randomisation number to a pharmacist managing the allocation sequence and the packaging of drugs at a central location.

Blinding of outcome assessment: the consulting physicians, pharmacist and women were blinded to allocation.

Documentation of exclusions: 107 women were lost to follow-up (however their pregnancy outcome was reported). Post randomisation 26 twins were excluded (multivitamin group: 15; iron/folic acid group: 11 twins (including one set of triplets). Only singleton pregnancies were included in the analysis because fetal loss and anthropometric measures at birth in multiple pregnancies are not primarily nutrition related. 3 women died before delivery and 1 woman underwent a therapeutic abortion.

Participants1374 women were recruited to participate, however 52 women were randomly assigned twice for consecutive pregnancies, resulting in data for 1426 pregnancies. Women had a pregnancy confirmed by urine testing and were randomly assigned to receive either IFA (n = 712) or UNIMMAP (n = 714) daily until 3 months after delivery. Women were recruited between 5 to 36 weeks’ gestation; 34.6% (n = 493) of the participants were recruited in the first trimester of pregnancy, mean gestational age at enrolment was 17.3 weeks (SD 7.8 wk).
Interventions

UNIMMAP: vitamin A 800 µg, Vitamin D 200 IU, Vitamin E 10 mg, Vitamin B-1 1.4 mg, Vitamin B-2 1.4 mg, Niacin 18 mg, Folic acid 400 µg, Vitamin B-6 1.9 mg, Vitamin B-12 2.6 µg, Vitamin C 70 mg, Zinc 15 mg, Iron 30 mg, Copper 2 mg, Selenium 65 µg, Iodine150 µg.

IFA (control): folic acid 400 µg, Iron 60 mg.

In a case of maternal illness, appropriate treatments were provided according to national guidelines. Severely anaemic women (haemoglobin < 70 g/L, without dyspnoea) received ferrous sulphate (200 mg) + folic acid (0.25 mg) twice daily, for 3 months, regardless of their allocation group. All participants also received 400 mg albendazole in the second and third trimesters. If malaria occurred despite chemoprophylaxis, quinine (300 mg, 3 times/day) was given for 5 days. Vitamin A (200,000 IU) was given to all women after delivery, in accordance with national recommendations.

Outcomes
  1. Gestational duration.

  2. Birthweight, birth length, and Rohrer ponderal index at birth (weight(g)X100/length3(cm)).

  3. Low birthweight (< 2500 g).

  4. Small-for-gestational age (birthweight below the 10th percentile).

  5. Large-for-gestational age (birthweight above the 90th percentile of the study population).

  6. Thoracic circumference, head circumference, mid upper arm circumference.

  7. Haemoglobin concentration in mothers during the third trimester, haemoglobin and sTfR concentrations in cord blood.

  8. Preterm birth (< 37 weeks’ gestation).

  9. Stillbirth (delivery of an infant showing no sign of life after a gestational age of 28 weeks).

  10. Perinatal death.

Notes

Women's risk of spontaneous and recurrent miscarriage was unclear. 18% of women in each group had experienced a previous fetal loss.

Women's nutritional status is unclear, although women are presumable at risk as the purpose of the trial is to correct multiple micronutrient deficiencies.

Intention to treat analyses not performed, however the review included details of losses to follow-up where the outcome was known.

Compliance: unclear, states that there was no difference in compliance between the two groups.

Location: Burkino Faso.

Timeframe: March 2004 to October 2006.

Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Low riskComputer generated with permuted blocks of 4.
Allocation concealment?Low riskRandomization numbers were kept in sealed opaque envelopes.
Blinding?
All outcomes
Low riskConsulting physicians, pharmacist and women were blinded to the intervention.
Incomplete outcome data addressed?
All outcomes
High riskData were reported for singletons only.
Free of selective reporting?Unclear riskAs above - data only reported for singletons.
Free of other bias?High riskSome women were pregnant more than once during the study period, however the denominators reported are the total number of pregnancies during the study period, not the total number of women randomised, which incorrectly assumes that each data point included is independent from the next.

Rumbold 2006

Methods

Randomisation and allocation concealment: computer generated random number list with balanced variable blocks and stratification for collaborating centre and gestational age (< 18 weeks vs 18 weeks or more), allocation occurred via a central telephone randomisation service. The treatment packs contained four sealed, opaque, white plastic bottles of either the antioxidants vitamin C and vitamin E or the placebo and were prepared by a researcher not involved in recruitment or clinical care.

Blinding of outcome assessment: women, caregivers and investigators were blinded to allocation.

Documentation of exclusion: no losses to follow-up.

Use of placebo control: placebo given.

Participants

1877 women were recruited into the study. Eligible women included those: with a nulliparous singleton pregnancy, between 14 and 22 weeks of gestation and with normal blood pressure at the first measurement in pregnancy and again at trial entry.

Women who had any of the following were excluded: known multiple pregnancy, known potentially lethal fetal anomaly, known thrombophilia, chronic renal failure, antihypertensive therapy, or specific contraindications to vitamin C or E therapy such as haemochromatosis or anticoagulant therapy.

Women were allocated to the vitamin C and E group (n = 935) or placebo (n = 935).

Interventions

Women allocated to the vitamin C and E group took four coated tablets of a combination of 250 mg of vitamin C (as ascorbic acid) and 100 IU of vitamin E (as d-alpha-tocopherol succinate) each day from trial entry until delivery (total daily dose of vitamin C: 1000 mg; vitamin E: 400 IU).

Women were advised not to take any other antioxidant supplements, although a multivitamin preparation that provided a daily intake of no more than 200 mg of vitamin C or 50 IU of vitamin E was permitted.

Outcomes
  1. Pre-eclampsia.

  2. A composite measure of death or serious outcomes in the infant.

  3. Small-for-gestational age.

  4. Serious infant complications occurring before hospital discharge.

  5. For women included a composite of any of the following until six weeks postpartum: death, pulmonary edema, eclampsia, stroke, thrombocytopenia, renal insufficiency, respiratory distress syndrome, cardiac arrest, respiratory arrest, placental abruption, abnormal liver function, preterm pre labor rupture of membranes, major postpartum haemorrhage, postpartum pyrexia, pneumonia, deep-vein thrombosis, or pulmonary embolus requiring anticoagulant therapy.

Notes

Women were at low risk of spontaneous and recurrent miscarriage based on the review criteria.

The majority of women participating had a baseline dietary intake of vitamin C and E above the Australian recommended daily amount.

Intention to treat analyses performed.

Compliance: There was no difference in compliance between the vitamin group (67%) and the placebo group (70%).

Location: Australia.

Timeframe: December 2001 and January 2005.

Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Low riskComputer generated random number list.
Allocation concealment?Low riskAllocation occurred via a central telephone randomisation service. Tablets were provided in sealed opaque bottles.
Blinding?
All outcomes
Low riskWomen, caregivers and investigators were blinded.
Incomplete outcome data addressed?
All outcomes
Low riskNo losses to follow-up.
Free of selective reporting?Low riskAll pre-specified outcomes reported.
Free of other bias?Low riskThe study appears to be free of other sources of bias.

Rumiris 2006

Methods

Generation of random number sequence: a computer generated random number sequence.

Randomisation and allocation concealment: central allocation (randomisation by an independent third party who had no conflict of interest in the study).

Blinding of outcome assessment: treatment allocations were blinded to both the investigator and the patient until the study was finished.

Documentation of exclusion: none reported.

Use of placebo control: no, comparisons were between antioxidants versus iron and folic acid.

Participants

60 women between 8 and 12 weeks gestation were eligible for randomisation (supplementation group: n = 29; folic acid group: n = 31).

Setting: at the antenatal clinic of the Department of Obstetrics and Gynecology, University of Indonesia between March 2003 and June 2004.

Eligibility criteria: pregnant women with low antioxidant status.

Exclusion criteria:

  1. history or current use of anti-hypertensive medication or diuretics;

  2. use of vitamins C >150 mg and/or E > 75 IU per day;

  3. known placental abnormalities;

  4. current pregnancy as a result of in vitro fertilisation;

  5. regular use of platelet active drugs or non-steroidal anti-inflammatory drugs (NSAIDs);

  6. known fetal abnormalities;

  7. documented uterine bleeding within a week of screening;

  8. uterine malformations;

  9. history of medical complications.

Interventions

Supplementation group: received antioxidant supplements daily - vitamins A (1000 IU), B6 (2.2 mg), B12 (2.2 ug), C (200 mg), E (400 IU), folic acid (400 ug), N-acetylcysteine (200 mg), Cu (2 mg), Zn (15 mg), Mn (0.5 mg), Fe (30 mg), calcium (800 mg), and selenium (100 ug).

Folic acid group: received Fe 30 mg and folic acid 400 ug daily.

Timing of the intervention: early pregnancy (8 to 12 weeks).

Outcomes
  1. Pre-eclampsia.

  2. Abortion.

  3. Hypertension.

  4. Intrauterine growth restriction.

  5. Intrauterine fetal death.

Notes

Women's risk of spontaneous and recurrent miscarriage was unclear.

Participating women had low antioxidant status at enrolment, as defined as superoxidedismutase level below 164U/mL. No nutritional information provided.

Intention to treat analyses performed.

Compliance: unclear, no information reported.

Location: Indonesia.

Timeframe: March 2003 and June 2004.

Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Low riskComputer generated random number sequence.
Allocation concealment?Low riskCentral allocation (randomisation by an independent third party who had no conflict of interest in the study).
Blinding?
All outcomes
Low riskTreatment allocations were blinded to both the investigator and the patient until the study was finished.
Incomplete outcome data addressed?
All outcomes
Low riskNo missing data.
Free of selective reporting?Low riskAll pre-specified outcomes were reported, no apparent evidence of selective reporting.
Free of other bias?Unclear riskAt baseline, the control group appears to have a 2 mmHg higher systolic blood pressure than the intervention group, this figure was of borderline statistical significance, P = 0.059.

Rush 1980

Methods

Randomisation and allocation concealment: unclear, women were allocated to groups based on "random assignment". Randomisation was stratified on pre-pregnancy weight, weight gain during pregnancy, previous low birthweight infant and protein intake. No other methodological details given.

Blinding of outcome assessment: unclear, women were allocated to 2 forms of treatment or control, where both treatments were given as a canned beverage and the control group were given standard oral multivitamins. No information is given on blinding of outcome assessors.

Documentation of exclusion: 237 women (22%) were excluded.

Use of placebo control: no placebo, the control group received standard prenatal multivitamin supplements.

Participants1051 women were recruited into the study. Women eligible were black, English speaking, and not greater than 30 weeks' gestation. They also had one of the following criteria: low pre-pregnant weight (under 110 pounds at conception); low weight gain at the time of recruitment; at least 1 previous low birthweight infant; a history of protein intake of less than 50 g in the 24 hours preceding recruitment. Women were not eligible if they were known to be seeking a termination, had specific chronic health disorders, if they admitted to recent use of narcotics or heavy use of alcohol, or weighed >= 140 pounds at conception.
The mean gestation at enrolment ranged from 16-18 weeks for the treatment groups.
1225 women were invited to join the study, of which 1051 (84%) consented. Of these, 237 (22%) were excluded and 814 women (77%) remained active in the study until delivery and were allocated to one of three groups: supplement (n = 263), complement (n = 272) or control (n = 279).
Interventions

Women were randomised to 1 of 3 groups:

  1. high protein supplement (daily 40 g animal protein, 470 calories, 1000 mg calcium, 100 mg magnesium, 60 mg iron, 4 mg zinc, 2 mg copper, 150 mcg iodine, 6000 IU vitamin A, 400 IU vitamin D, 30 USPU vitamin E, 60 mg vitamin C, 3 mg vitamin B1, 15 mg vitamin B2, 15 mg niacin, 2.5 mg vitamin B6, 1 mg pantothenic acid, 200 mcg biotin, 350 mcg folic acid, 8 mcg vitamin B12);

  2. balanced protein-energy complement (6 g animal protein, 250 mg calcium, 12 mg magnesium, 40 mg iron, 0.084 mg zinc, 0.15 mg copper, 100 mcg iodine, 4000 IU vitamin A, 400 IU vitamin D, 60 mg vitamin C, 3 mg vitamin B1, 15 mg vitamin B2, 10 mg niacin, 3 mg vitamin B6, 1 mg pantothenic acid, 350 mcg folic acid, 3 mcg vitamin B12);

  3. control (250 mg calcium, 0.15 mg magnesium, 117 mg iron, 0.85 mg zinc, 0.15 mg copper, 100 mcg iodine, 4000 IU vitamin A, 400 IU vitamin D, 60 mg vitamin C, 3 mg vitamin B1, 2 mg vitamin B2, 10 mg niacin, 3 mg vitamin B6, 1 mg pantothenic acid, 350 mcg folic acid, 3 mcg vitamin B12).

Women received the high protein or balanced protein-energy supplements in the format of a drink. Women in the control group received a standard oral prenatal multivitamin supplement.

Outcomes
  1. Total weight gain, average weight gain and early weight gain during pregnancy.

  2. Duration of gestation (presented as cumulative rates of delivery from life tables for each treatment group).

  3. Preterm birth < 37 weeks.

  4. Fetal death (before < 20 weeks' gestation and >= 20 weeks' gestation).

  5. Neonatal death (according to gestation at delivery).

  6. Birthweight (mean).

  7. Somatic measures of infant growth at 1 year of age.

  8. Psychological measures at 1 year of age.

NotesWomen's risk of spontaneous and recurrent miscarriage is unclear, as there is no information about concurrent medical conditions or other risk factors for miscarriage. Women in the trial had a low caloric intake at trial entry, and unexpectedly, an adequate protein intake. No other specific nutritional information is reported.
Sample-size calculation reported: 250 women were required in each treatment group to show a 125 g difference in birthweight. A 25% loss to follow-up was incorporated into the sample size.
Intention-to-treat analyses not performed.
There were 9 sets of twins amongst the three treatment groups.
Compliance: "on average, about three quarters of the prescribed amount of beverage was probably ingested".
Location: New York City, USA.
Timeframe: 1969 to 1976.
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Unclear riskNo methodological details given beyond reporting of 'random assignment'.
Allocation concealment?Unclear riskNo methodological details given beyond reporting of 'random assignment'.
Blinding?
All outcomes
Unclear riskUnlkely as women were given canned beverages or multivitamins.
Incomplete outcome data addressed?
All outcomes
Unclear risk237 women (22%) excluded, no intention to treat analysis.
Free of selective reporting?Unclear riskUnclear if all pre-specified outcomes reported.
Free of other bias?Unclear riskLimited methodological details provided.

Schmidt 2001

Methods

Randomisation and allocation concealment: unclear, women were "randomly assigned on an individual basis, to double-blind, weekly supplementation until delivery".

Blinding of outcome assessment: unclear, double blind stated in text but no details given.

Documentation of exclusion: 42 women (17%) were lost to follow-up and excluded.

Use of placebo control: control tablets containing iron and folic acid were given.

Participants243 women were recruited into this study. Pregnant women between 16 and 20 weeks' gestation, aged between 17 and 35 years old, with a parity < 6 and haemoglobin level between 80-140 g/l, were eligible for this study. Women were randomised to receive either vitamin A plus iron and folic acid (n = 122) or iron and folic acid only (n = 121). Of these 22 (18%) and 20 (17%) women in vitamin A plus iron and folic acid and the iron and folic acid groups respectively, dropped out between enrolment and the follow-up at 4 months.
InterventionsWomen were randomised to a weekly supplementation with 120 mg ferrous sulfate and 500 mcg folic acid, with or without vitamin A (2400 retinol equivalents). Women were asked to take the trial tablets from between 16 and 20 weeks' gestation until birth.
Outcomes1. Stillbirth.
2. Concentrations of haemoglobin, serum ferritin and serum transferrin receptors, at or near term.
3. Concentrations of iron and vitamin A in breast milk.
4. Haemoglobin and serum vitamin A concentrations in the mother and infant at 4 months postpartum.
5. General health, growth and development measures in the first year of life.
NotesWomen risk status for spontaneous and recurrent miscarriage is unclear.
At baseline, between 13% and 17% of women had marginal vitamin A deficiency 44% to 50% of women were anaemic.
Sample-size calculation performed allowing for a 50% drop-out during the study period.
Intention-to-treat analyses were not performed.
Compliance: adherence to the tablet intake was assessed through interview during a postnatal home visit, which revealed that the median tablet intake was 50 tablets (i.e. 25 weeks), while only 17% of the subjects took more than 90 tablets.
Location: Indonesia.
Serial publications of this data report different denominators.
Time frame: November 1997 to May 1998.
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Unclear riskNo information provided about sequence generation.
Allocation concealment?Unclear riskWomen were 'randomly assigned on an individual basis' but no other details given.
Blinding?
All outcomes
Unclear riskDouble blind used in the text but not details provided.
Incomplete outcome data addressed?
All outcomes
Unclear risk42 women (17%) were lost to follow-up and excluded, no intention to treat analyses performed.
Free of selective reporting?High riskSerial publications of this study report different denominators.
Free of other bias?Unclear riskLimited methodological details provided.

Spinnato 2007

Methods

Generation of random number sequence: the randomisation sequence was constructed by the data coordinating centre (DCC) as permuted blocks of random size, stratified by clinical centre, and implemented via a program residing on the clinical centres study computer.

Randomisation and allocation concealment: central allocation.

Blinding of outcome assessment: all clinicians and clinical investigators were blinded to group assignment.

Documentation of exclusion: none reported.

Use of placebo control: placebo control.

Participants

739 eligible women between 120/7 and 196/7 weeks of gestation were enrolled in the study (treatment: 371; placebo: 368).

Setting: four Brazilian clinical centres: one primary clinical centre (Recife) and 3 additional clinical sites (Campinas, Botucatu, and Porto Alegre); each site’s major teaching hospital serves a primarily urban low-income population.

Eligibility criteria: women between 120/7 and 196/7 weeks of gestation and diagnosed with nonproteinuric chronic hypertension or a prior history of pre-eclampsia in their most recent pregnancy that progressed beyond 20 weeks' gestation.

Exclusion criteria:multifetal gestation, allergy to vitamin C or vitamin E, requirement for aspirin or anticoagulant medication, 24-hour urinary protein ≥ 300 mg, pre-pregnancy diabetes mellitus, known fetal anomaly incompatible with life.

Loss to follow-up: 32 women (treatment 16; placebo 16).

Interventions

Intervention group: daily treatment with both vitamin C (1000 mg) and E (400 IU) until delivery or until the diagnosis of pre-eclampsia.

Control group: daily placebo until delivery or until the diagnosis of pre-eclampsia.

Timing of the intervention: between 120/7 and 196/7 weeks of gestation.

Outcomes
  1. Pre-eclampsia (women were followed through the 14th day postpartum for the occurrence of pre-eclampsia).

  2. Severity of pre-eclampsia.

  3. Gestational hypertension.

  4. Abruptio placentae.

  5. Premature rupture of membranes.

  6. Preterm birth.

  7. Small-for-gestational age.

  8. Low birthweight infant.

Notes

25 inclusion/exclusion criteria violations (23 enrolled outside 12-19 weeks’ gestation; 2 twin gestations - one lost to spontaneous abortions, one delivered liveborn in treatment group); all 25 women remained in their assigned study groups.

26 women had early treatment terminations (treatment 19; placebo 7), but remained in follow-up.

Women's risk of spontaneous and recurrent miscarriage was unclear.

Women's nutritional status is also unclear.

Intention to treat analyses performed.

Compliance: average compliance was 85%, and similar between treatment groups.

Location: Brazil.

Timeframe: July 2, 2003 and November 23, 2006.

Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Low riskComputer generated sequence number.
Allocation concealment?Low riskCentral allocation.
Blinding?
All outcomes
Low riskAll clinicians and clinical investigators were blinded to group assignment.
Incomplete outcome data addressed?
All outcomes
Low riskSmall numbers of missing data, balanced across groups (32 women; treatment 16; placebo 16).
Free of selective reporting?Low riskAll pre-specified outcomes were reported, no apparent evidence of selective reporting.
Free of other bias?Low riskThe study appears to be free of other sources of bias.

Steyn 2003

Methods

Randomisation and allocation concealment: "randomisation was undertaken by computer-generated numbers". Roche Pharmaceutical supplied numbered containers with either vitamin C or matching placebo, and they retained the study code until completion of the study. No other methodological details given.

Blinding of outcome assessment: "double blind" stated, Roche Pharmaceuticals retained the code until completion of the study.

Documentation of exclusion: none reported.

Use of placebo control: placebo control.

Participants200 women were recruited into the study. Women with a history of a previous mid-trimester abortion (defined as spontaneous expulsion of the uterine contents between 13 and 26 weeks' gestational age), or previous preterm labour, and less than 26 weeks' gestation were eligible and invited to participate. Women with iatrogenic causes of their previous preterm labour such as previous induction of labour before term for severe pre-eclampsia, were excluded. 203 consecutive women were approached, of which 200 (98.5%) consented and were randomised to either vitamin C (n = 100) or placebo (n = 100). No losses to follow-up were reported.
InterventionsTwice daily tablet of either 250 mg vitamin C or placebo, from trial entry until 34 weeks' gestation. All women were tested for bacterial vaginosis and all women with positive cultures for Mycoplasma hominis (and between 22 and 32 weeks' gestation) were treated with erythromycin for 7 days.
Outcomes
  1. Preterm labour, defined as spontaneous onset of labour and delivery before 37 completed weeks.

  2. The secondary outcome was perinatal outcome, a composite endpoint including birthweight, gestational age at delivery, perinatal mortality, duration of admission in the neonatal intensive care unit and neonatal complications.

The age of fetal viability was considered to be 28 weeks' gestation.

NotesResults are from an interim analysis performed when 100 participants were recruited into each arm. Recruitment was stopped after the interim analysis revealed few differences between the two groups. Unclear if there was a sample-size calculation performed. Women's risk profile spontaneous and recurrent miscarriage is unclear, although they are clearly at high risk of preterm birth. It is also unclear if multiple births were included.
6% of women had an inadequate dietary intake of vitamin C, defined as an intake < 67% of the recommended dietary allowance (70 mg per day).
Compliance: women were requested to bring the containers to each visit and the remaining tablets were counted to improve and control compliance; however, no compliance data were reported.
Country: South Africa.
Timeframe: unclear.
Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Low riskComputer generated sequence number.
Allocation concealment?Low riskRoche pharmaceuticals supplied numbered study containers and kept the study code until completion of the study.
Blinding?
All outcomes
Unclear riskDouble blind stated in the text but not details given.
Incomplete outcome data addressed?
All outcomes
Low riskNo losses to follow-up reported.
Free of selective reporting?Unclear riskRecruitment stopped after an interim analysis.
Free of other bias?High riskResults are from an interim analysis performed when 100 participants were recruited into each arm.

Taylor 1982

Methods

Generation of random number sequence: not reported.

Randomisation and allocation concealment: women were randomised, no further details given.

Blinding of outcome assessment: not reported.

Documentation of exclusion: no losses to follow-up reported, however three women were delivered before 37 weeks and were therefore excluded from the study.

Use of placebo control: no.

Participants

48 healthy pregnant women at 12 weeks' gestation (intervention group: 21; control group: 24).

Eligibility criteria: healthy pregnant women with no adverse medical or obstetric history.

Loss to follow-up: no.

Interventions

Intervention group: 325 mg of ferrous sulphate and 350 µg of folic acid to be taken daily throughout the remainder of pregnancy.

Control group: the women in non-therapy group were not given any supplements.

Timing of the intervention: from 12 weeks' gestation until delivery.

Outcomes
  1. Serum ferritin concentration.

  2. Mean cell volume.

  3. Haemoglobin concentration.

Notes

Women's risk of spontaneous and recurrent miscarriage was unclear.

Women's nutritional status is also unclear.

Intention to treat analyses: no, 3 women excluded.

Compliance: unclear, no information reported.

Location: Scotland.

Timeframe: unclear.

Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Unclear riskNot reported.
Allocation concealment?Unclear riskWomen were randomised, no further details given.
Blinding?
All outcomes
Unclear riskNot reported.
Incomplete outcome data addressed?
All outcomes
Low riskNo missing data; three women were delivered before 37 weeks and were therefore excluded from the study.
Free of selective reporting?Low riskAll pre-specified outcomes were reported, no apparent evidence of selective reporting.
Free of other bias?Unclear riskLimited methodological details provided.

The Summit 2008

Methods

Generation of random number sequence: computer-generated number; 262 clustered unit of randomisations (all pregnant women served by the same midwife received supplements with the same midwife identification number).

Randomisation and allocation concealment: central allocation.

Blinding of outcome assessment: all study scientists and personnel, government staff, and enrollees were unaware of the allocation.

Documentation of exclusion: 1748 loss to follow-up before delivery (IFA: 853; MMN: 895); 1128 loss to follow-up after delivery (IFA: 553; MMN: 575). 10,549 pregnant women excluded post-randomisation because of trial termination (IFA group: 5057; MMN group: 5492).

Use of placebo control: no placebo, comparisons were between multiple micronutrients and iron and folic acid.

Participants41,839 pregnant women of any gestational age living on Lombok, Nusa Tenggara Barat Province, Indonesia. Women were allocated to iron and folic acid (n = 20,543) or multiple micronutrient (n = 21,296).
Interventions

MMN group: the MMN was the UNIMMAP formulation containing 30 mg iron (ferrous fumarate) and 400 ug folic acid along with 800 ug retinol (retinyl acetate), 200 IU vitamin D (ergocalciferol), 10 mg vitamin E (alpha tocopherol acetate), 70 mg ascorbic acid, 1.4 mg vitamin B1 (thiamine mononitrate), 18 mg niacin (niacinanide), 1.9 mg vitamin B6 (pyridoxine), 2.6 ug vitamin B12 (cyanocobalamin), 15 mg zinc (zinc gluconate), 2 mg copper, 65 ug selenium, and 150 ug iodine - one capsule daily up to 3 months after birth.

IFA group: the IFA contained 30 mg iron (ferrous fumarate) and 400 ug folic acid - one capsule daily up to 3 months after birth.

Timing of the intervention: any time during pregnancy.

Outcomes
  1. Early infant mortality (deaths until 90 days postpartum).

  2. Neonatal mortality.

  3. Fetal loss (abortions and stillbirths).

  4. Low birthweight.

Notes

Women's risk of spontaneous and recurrent miscarriage was unclear.

Women's nutritional status is also unclear. However, 30% of women in each group had an mid upper arm circumference < 23.5cm, which was used as an indicator of women being undernourished.

Intention to treat analyses performed.

Compliance: median compliance was 85%, there was no difference between treatment groups in compliance.

Location: Indonesia.

Timeframe: July 1, 2001, and April 1, 2004.

Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Low riskComputer generated number.
Allocation concealment?Low riskCentral allocation.
Blinding?
All outcomes
Low riskAll study scientists and personnel, government staff, and enrollees were unaware of the allocation.
Incomplete outcome data addressed?
All outcomes
High risk

Loss to follow-up: 1748 loss to follow-up before delivery (IFA: 853; MMN: 895); 1128 loss to follow-up after delivery (IFA: 553; MMN: 575).

Post-randomisation exclusion: 10,549 pregnant women excluded because of trial termination (IFA group: 5057; MMN group: 5492).

Free of selective reporting?Low riskAll pre-specified outcomes were reported, no apparent evidence of selective reporting.
Free of other bias?Low riskThe study appears to be free of other sources of bias.

Van den Broek 2006

Methods

Generation of random number sequence: using a random-generation procedure.

Randomisation and allocation concealment: the supplements in vitamin A and placebo treatments allocated were prepared in identical capsules and were packaged in bottles according to the randomisation schedule (sealed envelopes) by midwives who were not involved in the trial conduct.

Blinding of outcome assessment: neither the women nor the midwives involved in treatment allocation revealed the randomisation schedule to anyone involved in the conduct of the trial.

Documentation of exclusion: 77 loss to follow-up before assessment at 26-28 weeks (5000 IU vitamin A: 26; 10,000 IU vitamin A: 26; placebo: 25). Additional 93 loss to follow-up before assessment at 36-38 weeks (5000 IU vitamin A: 34; 10,000 IU vitamin A: 28; placebo: 31).

Use of placebo control: placebo control.

Participants

Seven hundred women with singleton pregnancies at 12-24 weeks gestation measured by ultrasound scan (5000 IU vitamin A: 234; 10,000 IU vitamin A: 234; placebo: 232).

Setting: the antenatal clinic at the Namitambo rural Health Centre in southern Malawi, central Africa.

Eligibility criteria: (Hb) < 11.0 g/dl by HemoCue screening method at first antenatal visit, singleton pregnancy with gestational age > 12 weeks and ≤ 24 weeks measured by ultrasound scan, no fetal abnormality detectable by ultrasound at time of booking, residing in the catchment area of the health centre.

Exclusion criteria: women > 24 weeks' gestation, or twin pregnancy.

Interventions
  • Intervention group 1: 5000 IU vitamin A daily until delivery.

  • Intervention group 2: 10,000 IU vitamin A daily until delivery.

  • Comparison group: placebo daily until delivery.

Timing of the intervention: supplementation started as early as possible after 12 weeks of pregnancy.

All women received iron tablets daily (60 mg elemental iron as ferrous sulphate with 0.25 mg folic acid).

Outcomes
  1. Anaemia status (no anaemia ([Hb]≥11.0 g/dl), anaemia ([Hb] < 11.0 g/dl) or severe anaemia ([Hb] < 8.0 g/dl).

  2. Haemoglobin concentration (Coulter counter value), iron status (determined by serum ferritin and serum transferring receptor concentration).

  3. Evidence of infection (assessed by serum CRP, peripheral malaria parasitaemia and HIV status).

  4. Vitamin A status (determined by serum retinol and the MRDR).

Notes

Women's risk of spontaneous and recurrent miscarriage was unclear.

Women's nutritional status is also unclear.

Intention to treat analyses performed.

Compliance: unclear, no information provided.

Location: Malawi.

Timeframe: April 1997 and July 1999.

Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Low riskRandom-generation procedure used.
Allocation concealment?Low riskThe vitamin A and placebo treatments allocated were prepared in identical capsules and packaged in bottles according to the randomisation schedule (sealed envelopes) by midwives who were not involved in the trial conduct.
Blinding?
All outcomes
Low riskNeither the women nor the midwives involved in treatment allocation revealed the randomisation schedule to anyone involved in the conduct of the trial.
Incomplete outcome data addressed?
All outcomes
High risk77 loss to follow-up before assessment at 26-28 weeks (5000 IU vitamin A: 26; 10,000 IU vitamin A: 26; placebo: 25). Additional 93 loss to follow-up before assessment at 36-38 weeks (5000 IU vitamin A: 34; 10,000 IU vitamin A: 28; placebo: 31).
Free of selective reporting?Low riskAll pre-specified outcomes were reported, no apparent evidence of selective reporting.
Free of other bias?Low riskThe study appears to be free of other sources of bias.

Villar 2009

  1. a

    d: day
    F: folic acid
    HbCC: haemoglobin C disease
    HbSc:haemoglobin SC disease
    HbSS: haemoglobin sickle cell disease
    HELLP syndrome: haemolysis, elevated liver enzymes, low platelet count syndrome
    HIV-1: Human Immunodeficiency Virus-1
    HOFPP: Hungarian Optimal Family Planning Programme
    IQR: interquartile range
    IFA: iron and folic acid
    IU: international units
    IVF-ET: in vitro fertilization and embryo transfer
    mcg: micrograms
    mg/mL: milligrams per millilitre
    MF: multivitamins with folic acid
    mg: milligrams
    MMN: multiple micronutrient
    MRDR: modified relative dose-response
    MV: multivitamins without folic acid
    MRC: Medical Research Council
    NTD: neural tube defect
    P: progesterone
    PAI-1: plasminogen activator inhibitor-1
    PAI-2: plasminogen activator inhibitor-2
    PCV: packed cell volume
    UK: United Kingdom
    UNIMMAP: United Nations International Multiple Micronutrient Preparation
    USA: United States of America
    WBC: white blood cell
    wk: week

Methods

Generation of random number sequence: no sequence generation details available.

Randomisation and allocation concealment: central allocation (randomisation was performed by the statisticians of the British VIP Trial).

Blinding of outcome assessment: "double blind" stated.

Documentation of exclusion: 10 women (treatment 6; placebo 4), and 29 infants (treatment 13, placebo 16) were lost to follow-up.

Use of placebo control: placebo control.

Participants

1365 women between14-22 gestational age agreed to participate and were randomised (vitamins group: 687; placebo group: 678).

Setting: antenatal clinics located in Nagpur, India; Lima and Trujillo, Peru; Cape Town, South Africa; and Ho Chi Minh City, Viet Nam which served populations with low social-economic status and had evidence of overall low nutritional status, between October 2004 and December 2006.

Eligibility criteria: pregnant women considered high risk for pre-eclampsia (chronic hypertension, renal disease, pre-eclampsia-eclampsia in the pregnancy preceding the index pregnancy requiring delivery before 37 weeks’ gestation, HELLP syndrome in any previous pregnancy, pre-gestational diabetes, primiparous with a body mass index > 30 kg/m2, history of medically indicated preterm delivery, abnormal uterine artery Doppler waveforms and women with antiphospholipid syndrome), multifetal gestation. Women ingesting medications with aspirin-like compounds were not excluded.

Exclusion criteria: women ingesting vitamin supplements that contained ≥ 200 mg of vitamin C and/or ≥ 50 IU of vitamin E and women receiving warfarin.

Interventions

Intervention group: received 1000 mg vitamin C and 400 IU of vitamin E daily until delivery.

Comparison group: received placebo daily until delivery.

Timing of the intervention: between 14 and 22 weeks' gestation.

Outcomes
  1. Pre-eclampsia.

  2. Eclampsia.

  3. Placental abruption.

  4. Low birthweight (LBW) (< 2500 g).

  5. Small-for-gestational age (< 10th centile of the WHO recommended standard).

  6. Intrauterine or neonatal death before hospital discharge.

  7. Preterm delivery (< 37 weeks).

  8. Early preterm delivery (< 34 weeks).

  9. Very LBW (< 1500 g).

  10. ≥ 7 days in the neonatal intensive care unit.

  11. Congenital malformations.

Pre-eclampsia information was unavailable for 14 women in the vitamins and 9 in the placebo group.

There were data from 81 supplemented (11.8%) and 100 placebo-treated (14.7%) women with multiple pregnancies, for whom newborn outcomes were considered separately.

Notes

Women's risk of spontaneous and recurrent miscarriage was unclear. Women at high risk of pre-eclampsia were included but data on fetal loss was not reported separately for this group.

No specific information on women's nutritional status is included; however, the paper states that the trial was conducted in populations with 'documented low nutritional status'.

Intention to treat analyses performed.

Compliance: Median compliance was 87%, and was similar between the treatment groups.

Location: Antenatal clinics in India, Peru, South Africa and Viet Nam.

Timeframe: October 2004 and December 2006.

Risk of bias
BiasAuthors' judgementSupport for judgement
Adequate sequence generation?Low riskRandomisation sequence blocked by centre in groups of 2 to ten10 individuals.
Allocation concealment?Low riskCentral allocation (randomisation was performed by the statisticians of the British VIP Trial).
Blinding?
All outcomes
Low riskWomen and investigators blinded to allocation.
Incomplete outcome data addressed?
All outcomes
Low risk

Small numbers of missing data, balanced across groups.

Women: 10 (treatment 6; placebo 4).

Infants: 29 (treatment 13; placebo 16).

Free of selective reporting?Unclear riskPerinatal death was reported instead of pre-specified neonatal death.
Free of other bias?Low riskThe study appears to be free of other sources of bias.

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
Baumslag 1970Onset of supplementation was > 20 weeks' gestation.
Women were supplemented with either iron, iron and folic acid or iron, folic acid and vitamin B12 from "after the 24th week of pregnancy".
Biswas 1984Unclear of the gestational age at which women entered the trial.
Blot 1981Onset of supplementation was > 20 weeks' gestation.
Supplementation with either iron and folic acid or iron alone occurred "at the end of the 6th month of pregnancy". Unclear if women were randomised to the treatment groups.
Chanarin 1968Onset of supplementation was > 20 weeks' gestation.
Women were given a folic acid supplement after the 20th week of pregnancy. Abortion was reported according to folic acid status at 15 weeks, prior to supplementation.
Colman 1974Onset of supplementation was > 20 weeks' gestation.
Women were supplemented "during the final month of pregnancy". Outcomes reported included folic acid red cell and serum folic acid concentration and haemoglobin concentration.
Coutsoudis 1999Onset of supplementation was > 20 weeks' gestation.
Women were given vitamin A and beta-carotene "during the third trimester of pregnancy".
Dawson 1962Onset of supplementation was > 20 weeks' gestation.
Women were supplemented with folic acid "on or after the 28th week". Group allocation was not done randomly. Reported outcomes include incidence of folic acid deficiency and megaloblastic anaemia, and haemoglobin concentration.
Edelstein 1968Onset of supplementation was > 20 weeks' gestation.
Supplementation was started at the 28th week of pregnancy. Outcomes reported included serum folate activity and serum folate, urinary formiminoglutamic acid, serum vitamin B12, mean haemoglobin and haematocrit values.
Ferguson 1955Only 24 (9%) of the 269 women in the trial began to participate before 15 weeks' gestation and outcomes not reported separately according to gestation at enrolment.
Feyi-Waboso 2005Onset of supplementation was 20 or more weeks' gestation.
Fletcher 1971No inclusion/exclusion criteria reported, unclear of gestational age at enrolment to the study, reports combined outcomes for "antepartum and threatened or complete abortion" and "stillbirth or neonatal death or congenital malformation" (not reported separately).
Giles 1971Onset of supplementation was > 20 weeks' gestation for a large proportion of the participants.
4 groups in the study, two of which involved supplementation after 20 weeks' gestation. Results were not reported separately between groups.
Hampel 1974Unclear of the gestational age at which women entered the trial.
Hankin 1966No main outcomes reported.
Supplementation was from "approximately 20 weeks", no clinically relevant outcomes, outcomes relating to vitamin C status in plasma and breast milk reported.
Hibbard 1968No main outcomes reported.
Biochemical measures of blood folate status reported.
Hunt 1984All women received a multivitamin in addition to the zinc supplement or placebo.
Huybregts 2009Both groups received a multivitamin supplement (same vitamin content in each group).
Laurence 1981No main outcomes or pregnancy loss outcomes reported. Miscarriage reported in those women where there was a neural tube defect, but not in all women according to treatment group.
Lira 1989No main outcomes reported.
Biochemical measures of iron and folate status reported.
Lumeng 1976Unclear gestational age at enrolment, 5 women were excluded due to abortion, premature labour, inadequate dietary records or missing more than 3 prenatal visits. Exclusions were not reported by group allocation. Outcomes related to maternal and fetal plasma levels of pyridoxal 5'-phosphate and coenzyme saturation of aspartate aminotransferase and alanine aminotransferase in maternal erthrocytes were reported.
Marya 1981Onset of supplementation was > 20 weeks' gestation.
Women were supplemented with vitamin D "throughout the 3rd trimester".
Meirinho 1987No clinical outcomes reported.
Maternal plasma concentrations of trophoblastic protein SP1 were reported.
Metz 1965Onset of supplementation was > 20 weeks' gestation.
Women were supplemented with either iron or iron and folic acid, or iron, folic acid and vitamin B12. Supplementation was started after the 24th week of pregnancy.
Mock 2002No main outcome reported.
Women were enrolled in either early or late pregnancy. Biochemical measures of biotin status reported.
Moldenhauer 2002No main outcomes reported.
Unclear if this is a cohort study or randomised trial. Women in this study were participating in a randomised placebo controlled trial of calcium supplementation, and completed a dietary assessment at 12-21 weeks' gestation and 29-31 weeks' gestation. Unclear whether all women took a standard prenatal multivitamin or just women in the placebo group. Results are presented according to "teens", "twins" and "singleton" pregnancies, not according to whether women took the supplement or not. Outcomes reported included dietary intakes of vitamin C and E (with and without the contribution of the prenatal vitamin supplement).
Owen 1966Onset of supplementation was > 20 weeks' gestation.
Women supplemented with oral vitamin K1 "several days before delivery".
Ross 1985Unclear about content of vitamin supplements.
Women were supplemented with high or low 'bulk' dietary supplements with vitamins added; however, the vitamin supplements added were not specified.
Schuster 1984Unclear of gestation at enrolment to the trial.
No pregnancy loss outcomes reported.
Semba 2001No main outcomes reported.
Women enrolled between 18 and 28 weeks' gestation, no clinical outcomes reported, only haemoglobin and plasma erythropoietin concentrations.
Shu 2002Both groups received a multivitamin (same vitamin content in both groups).
Smithells 1981Non-randomised study of periconceptional multivitamin supplementation for the prevention of neural tube defects.
Suharno 1993No main outcomes reported.
Anaemic pregnant women were enrolled between 16 and 24 weeks' gestation. The only clinical outcome reported was the percentage of women with anaemia following treatment with a combination of vitamin A and iron or placebo.
Tanumihardjo 2002No main outcomes reported.
Mean gestation at enrolment was 17.6 weeks, no clinical outcomes reported, markers of vitamin A and iron status reported.
Thauvin 1992No main outcomes reported.
Women were supplemented from 3 months' gestation, data on pregnancy outcomes including spontaneous abortion were collected but not reported.
Trigg 1976Unclear of gestation at enrolment to the trial.
Ulrich 1999Non-randomised study.
Observational cohort study of folic acid users, randomised to different doses of folic acid, but no controls.
Villamor 2002No main outcomes reported.
Women enrolled between 12 and 27 weeks' gestation, no pregnancy loss or main outcomes reported, only reports measures of weight gain during pregnancy.
Vutyavanich 1995No main outcomes reported.
Women were enrolled in the study if they were less than 17 weeks' gestation; however, no pregnancy loss or main outcomes were reported, only measures of nausea and vomiting.

Characteristics of studies awaiting assessment [ordered by study ID]

Chelchowska 2004

MethodsUnlcear.
Participants138 pregnant women recruited from the Mother and Child Institute Hospital, Warsaw, Poland. Inclusion criteria included: maternal good health, no smoking, normal pregnancy, no vitamin and mineral supplementation prior to 12 weeks' gestation, and no fetal development defects.
InterventionsWomen took either the VIBOVIT®mama preparation or a placebo from 12 weeks' gestation until delivery. The VIBOVIT®mama preparation consisted of: vitamin D (400 IU), vitamin A (2000 IU), beta-carotene (3000 mcg), vitamin E (18 IU), zinc (15 mg), copper (2 mg) and selenium (20 mcg).
OutcomesLipid peroxidation and activity of superoxide dismutase and selenium-dependent glutathione peroxidase in blood samples taken from the mother and the cord blood of the infant.
NotesTranslated from Polish. Appears to be the same study as the Kubik 2004 paper, but reporting biochemical outcomes.

Frenzel 1956

MethodsUnclear.
ParticipantsUnclear.
InterventionsUnclear.
OutcomesUnclear.
NotesA copy of the paper could not be located.

Kubik 2004

MethodsUnclear, described in the abstract as: "healthy pregnant women were divided by a double blinded trial into a test group taking vitamin and mineral supplementation and a control group taking placebo".
Participants138 pregnant women recruited from the Mother and Child Institute Hospital, Warsaw, Poland. Inclusion criteria included: maternal good health, no smoking, normal pregnancy, no vitamin and mineral supplementation prior to 12 weeks' gestation, and no fetal development defects.
InterventionsWomen took either the VIBOVIT®mama preparation or a placebo from 12 weeks' gestation until delivery. States in the abstract that women took a vitamin and mineral supplement (VIBOVIT®mama) which contained zinc (15 mg), copper (2 mg) and selenium (20 mcg).
OutcomesPregnancy induced hypertension, mode of birth, birthweight, Apgar scores, and other outcomes related to the 'course of pregnancy and delivery', biochemical measures of antioxidant status.
NotesTranslated from Polish. Actual numbers are not presentation, only %, and it is unclear how many women were allocated to each group at the onset of the study.

Characteristics of ongoing studies [ordered by study ID]

Fall 2007

Trial name or titleMumbai Maternal Diet Study: randomised controlled trial of micronutrient-dense food before and during pregnancy to prevent low birthweight.
MethodsRandomised controlled trial.
Participants

Inclusion criteria:

  1. women living in slum communities in Bandra and Khar districts of Mumbai served by the Women of India Network (WIN) primary health care clinics;

  2. women who wish to join;

  3. married;

  4. aged 15 to 35 years;

  5. not pregnant at recruitment;

  6. not using any PERMANENT form of contraception;

  7. intending to have more children;

  8. planning any future deliveries in Mumbai.

Exclusion criteria:

  1. women living outside the study area;

  2. non-married women;

  3. women outside the age range specified;

  4. women currently pregnant (these may become eligible after delivery);

  5. women who have undergone sterilisation surgery, or whose husbands have had a vasectomy;

  6. women definitely not planning further pregnancies;

  7. women definitely planning further deliveries outside Mumbai.

InterventionsWomen who are not pregnant, but are planning to have further children, will be recruited and randomised to one of four groups, to receive one of two interventions: a daily food-based supplement made from vegetables, fruit, and milk, of differing micronutrient content. Supplementation will be supervised. Field staff will record menstrual dates, in order to detect pregnancy as early as possible. Women who become pregnant will have investigations during pregnancy, including blood samples and ultrasound scans.
Outcomes
  1. Birthweight.

  2. Infant mortality.

  3. Maternal micronutrient status.

  4. Maternal infection load and immune status.

  5. Fetal losses (miscarriages and stillbirths).

  6. Newborn body composition.

  7. Newborn immune function.

Starting date09/01/2006.
Contact information

Dr Caroline Fall

MRC Epidemiology Resource Centre
Southampton General Hospital
University of Southampton
Tremona Road

NotesAnticipated end date: 31/03/2010, listed as completed.

Johns 2004

Trial name or titleThe effect of antioxidant supplementation on women with threatened miscarriage.
MethodsRandomised controlled trial.
Participants580 women who present with first trimester bleeding.
InterventionsVitamin C 1000 mg and Vitamin E 400 IU
versus placebo.
Outcomes
  1. Incidence of miscarriage.

  2. Late miscarriage.

  3. Pre-term labour.

  4. Pre-term pre-labour rupture of the membranes.

  5. Fetal growth restriction.

  6. Pre-eclampsia.

Starting date01/03/2004.
Contact information

Dr  Jemma  Johns

UCLH/UCL Research & Development Governance Committee
Research and Development Directorate
University College London Hospitals NHS Trust
1st Floor, Maple House
149 Tottenham Court Road

NotesListed as completed.

Sezikawa 2007

Trial name or titleVitamin C and E Supplementation in Pregnant Women With Low Antioxidant Status.
MethodsRandomised controlled trial.
Participants

Pregnant women with low antioxidant status at 10-12 weeks gestation age.

Inclusion criteria:

  • agree to consent form, and consent to protocol of research;

  • Known healthy singleton 6-10 weeks pregnant women.

Exclusion criteria:

  • blood pressure > 135/85;

  • proteinuria;

  • history or current use of anti-hypertensive medication or diuretics;

  • use of vitamins C > 150 mg and/or E > 75 IU per day;

  • pregestational diabetes;

  • known placental abnormalities;

  • current pregnancy is a result of in vitro fertilisation;

  • regular use of platelet active drugs or non-steroidal anti-inflammatory drugs;

  • known fetal abnormalities;

  • documented uterine bleeding within a week of screening;

  • uterine malformations;

  • history of medical complications;

  • illicit drug or alcohol abuse during current pregnancy;

  • intent to deliver elsewhere;

  • known psychologic problems;

  • participating in another interventional study.

InterventionsVitamin C 1000 mg and E 400 IU versus placebo.
Outcomes
  1. Pre-eclampsia.

  2. Other adverse pregnancy outcomes.

Starting dateOctober 2006.
Contact information

Akihiko Sekizawa, MD, PhD

Showa University School of Medicine

NotesListed as completed, last updated February 16th, 2010.