1. Objectives

To evaluate the effects of vitamin E supplementation during pregnancy on pregnancy outcomes, adverse events, side effects and the use of health services

2. How studies were identified

The following databases were searched in March 2015:

• Cochrane Pregnancy and Childbirth Group’s Trials Register
• CENTRAL (The Cochrane Library 2015)
• MEDLINE
• Embase
• CINAHL

Relevant journals, conference proceedings and reference lists were also searched

3. Criteria for including studies in the review

3.1 Study type

Randomized controlled trials, including quasi-randomized trials

3.2 Study participants

Pregnant women living in areas where dietary vitamin E intakes were considered either adequate or inadequate

3.3 Interventions

Vitamin E supplementation, alone or in combination with other supplements, compared with placebo or no supplementation

3.4 Primary outcomes

Maternal

• Clinical preeclampsia
• Haematological measures: haemolytic anaemia, reticulocytosis, hyperbilirubinaemia, haemoglobin concentrations
• Preterm birth (<37 weeks’ gestation)

Neonatal

• Stillbirth, neonatal death, perinatal death
• Haematological measures: haemolytic anaemia, reticulocytosis, hyperbilirubinaemia, haemoglobin concentrations
• Intrauterine growth restriction (birth weight <3^rd centile or the most extreme centile reported)

Secondary outcomes for the mother included prelabour rupture of membranes (PROM, both preterm and term), death up to six weeks postpartum, elective delivery (induction of labour or elective caesarean section), caesarean section, bleeding episodes (placental abruption, antepartum haemorrhage, postpartum haemorrhage, complications of epidural anaesthesia, need for transfusion), measures of serious maternal morbidity (eclampsia, liver failure, renal failure, disseminated intravascular coagulation, pulmonary oedema, peripheral neuropathy), adverse events of vitamin E supplementation sufficient to cease supplementation, side effects of vitamin E supplementation (fatigue, weakness, altered coagulation times, immunosuppression, creatinuria, dermatitis, altered thyroid function, increased urinary androgen excretion), maternal satisfaction with care, antenatal hospital admission, visits to day care units, use of intensive care, ventilation, and dialysis. Secondary outcomes for the neonate included birth weight, infant death, gestational age at birth, congenital malformations, Apgar score less than seven at five minutes, vitamin K deficiency bleeding or haemorrhagic disease of the newborn, respiratory distress syndrome, chronic lung disease or bronchopulmonary asthma, periventricular haemorrhage, bacterial sepsis, necrotising enterocolitis, retinopathy of prematurity, peripheral neuropathy, disability at childhood follow-up (cerebral palsy, intellectual disability, hearing or visual impairment), poor childhood growth, admission to neonatal intensive care unit, duration of mechanical ventilation, length of stay in hospital, development, and special needs after discharge.

4. Main results

4.1 Included studies

Twenty-one randomized controlled trials, enrolling 22,129 women, were included in this review

• Nine trials recruited women at high or increased risk of preeclampsia and three trials included women with established preeclampsia; two trials involved women with established preterm PROM; one trial enrolled pregnant women with leg cramps; and one trial enrolled women with planned caesarean section after 35 weeks’ gestation
• Supplementation was initiated in the second trimester in most trials; however, the timing of initiation ranged from eight weeks’ gestation to 35 weeks’ or more
• In three trials vitamin E was provided alone; in 17 trials vitamin E was given in combination with vitamin C, and in two of these trials another intervention was included (aspirin and fish oil or allopurinol); and in one study women were supplemented with vitamin E-rich palm oil
• Most trials (15/21 trials) provided 400 IU of vitamin E per day; one trial each provided 100 mg, 200 international units (IU) or 800 IU of vitamin E per day; and in three trials the dose was unknown

4.2 Study settings

• Australia, Brazil, India (3 trials), Iran (3 trials), Malaysia, South Africa (2 trials), Turkey, the United Kingdom of Great Britain and Northern Ireland (2 trials), the United States of America (3 trials), Venezuela, and multi-site (3 trials: India, Peru, South Africa, and Viet Nam; Canada and Mexico; and the United Kingdom of Great Britain and Northern Ireland and the Netherlands)
• One multi-site trial was undertaken in populations of “low nutritional status” (low socioeconomic populations in India, Peru, South Africa, and Viet Nam), one trial was conducted in women with low intakes of vitamin E (42 to 43% of the recommended daily intake), and one trial reported women as having adequate vitamin E intakes. One trial reported preeclampsia outcomes according to women’s baseline vitamin E status

4.3 Study settings

How the data were analysed
One comparison was made: vitamin E in combination with other supplements versus placebo or no supplementation. Dichotomous data were summarized using risk ratios (RR) and 95% confidence intervals (CI), and continuous data were summarized using mean differences (MD) and 95% CI. When 10 or more trials were included in a meta-analysis, funnel plot asymmetry was assessed for publication bias. Sensitivity analyses were conducted by excluding studies with overall high or unclear risk of bias. Where substantial clinical or statistical heterogeneity (I²>30%) was present, random effects models were used. To explore potential sources of heterogeneity, the following subgroup analyses were planned for primary outcomes:

• Dosage of vitamin E: ≥7 mg alpha-tocopherol equivalents versus <7 mg alpha-tocopherol equivalents
• Gestational age at trial entry: <20 weeks versus ≥20 weeks
• Dietary adequacy of vitamin E intakes: low versus adequate
• Vitamin E alone versus in combination with other micronutrients
• Risk of adverse pregnancy outcomes at study entry (as defined by trialists)

Results
Vitamin E in combination with other supplements versus placebo or no supplementation
Primary outcomes
No statistically significant difference between vitamin E and control groups was found for the risk of preterm birth (RR 0.98, 95% CI [0.88 to 1.09], 11 trials/20,565 women) or clinical preeclampsia (RR 0.91, 95% CI [0.79 to 1.06], 14 trials/20,878 participants). No evidence of a difference between groups was found for the infant primary outcomes stillbirth (RR 1.17, 95% CI [0.88 to 1.59], 9 trials/19,023 participants), neonatal death (RR 0.81, 95% CI [0.58 to 1.13], 9 trials/18,617 participants), infant death (RR 3.02, 95% CI [0.12 to 74.12], 1 trial/2694 participants), infant hyperbilirubinaemia (RR 0.78, 95% CI [0.59 to 1.04], 1 trial/725 participants), perinatal death (RR 1.09, 95% CI [0.77 to 1.54], 6 trials/16,923 participants), or intrauterine growth restriction (RR 0.98, 95% CI [0.91 to 1.06], 11 trials/20,202 participants). No trials reported on the primary outcomes haemolytic anaemia, reticulocytosis, or haemoglobin concentrations for either mother or infant.

Assessment of reporting bias, sensitivity analyses and subgroup analyses for primary outcomes
Funnel plots for preterm birth and intrauterine growth restriction did not show any evidence of publication bias, while the funnel plot for clinical preeclampsia was asymmetric. Results did not alter meaningfully in sensitivity analyses excluding trials at risk of bias, or in subgroup analyses by risk of adverse pregnancy outcome. In subgroup analysis by gestation at trial entry, a reduced risk of clinical preeclampsia was found among trials with gestation unknown (RR 0.32, 95% CI [0.13 to 0.79], p=0.013; 2 trials/693 women; p=0.03 for subgroup differences). The risk of clinical preeclampsia was also reduced in subgroup analysis by dietary vitamin E intake, with pooled analysis of 10 trials in which dietary intake was unclear showing a statistically significantly 26% reduction in risk (RR 0.74, 95% CI [0.56 to 0.98], p=0.036; 6928 women), and analysis of women with a low baseline vitamin E status demonstrated a borderline significant 75% reduction in risk (RR 0.35, 95% CI [0.12 to 1.02], p=0.055; 1 trial/95 women; p=0.04 for subgroup differences). Subgroup analyses by dosage of vitamin E were not conducted as all trials provided more than 7 mg alpha-tocopherol equivalents per day, and subgroup analyses of vitamin E provided alone could not be performed due to a lack of data.

Additional outcomes
The risk of placental abruption was reduced with vitamin E supplementation in comparison with control (RR 0.64, 95% CI [0.44 to 0.93], p=0.02; 7 trials/14,922 women). The risk of term PROM was increased among women receiving vitamin E supplementation (RR 1.77, 95% CI [1.37 to 2.28], p=0.000012; 2 trials/2504 participants), but no significant effect of supplementation was found on preterm PROM (RR 1.27, 95% CI [0.93 to 1.75], 5 trials/1999 participants). No clear effect of vitamin E was found for other maternal secondary outcomes, including antepartum haemorrhage, maternal death, any caesarean section, prelabour caesarean section, induction of labour, eclampsia, renal failure or renal insufficiency, disseminated intravascular coagulation, pulmonary oedema, admission to adult intensive care unit, and hospitalizations during pregnancy. No significant effect of vitamin E supplementation was found for the infant outcomes mean birth weight, gestational age at birth, bacterial sepsis, necrotising enterocolitis, chronic lung disease/bronchopulmonary dysplasia, congenital malformations, Apgar score less than seven at five minutes, respiratory distress syndrome, periventricular haemorrhage and periventricular leukomalacia, retinopathy of prematurity, admission to neonatal intensive care unit, or use of mechanical ventilation. No other pre-specified secondary outcomes were reported on.

Adverse events and side effects
None of the included trials reported on adverse events sufficient to cease vitamin E supplementation. In one trial involving 1877 participants, the risk of abdominal pain was increased in the treatment group (RR 1.66, 95% CI [1.16 to 2.37]). No evidence of a difference between treatment and control groups was found for other adverse outcomes, including elevated liver enzymes, acne, transient weakness, skin rash, or any side effect.

5. Additional author observations*

The overall methodological quality of the included trials was fair, with several of the largest trials considered to be at low risk of bias contributing much of the data to analyses. Using the GRADE approach, evidence for the outcomes preterm birth, intrauterine growth restriction and placental abruption was rated as high quality, evidence for the outcomes stillbirth and clinical preeclampsia was rated as moderate quality, and for preterm PROM the evidence was regarded as low quality. Data from the three identified trials in which vitamin E was given alone could not be included in analyses, and thus treatment effects may not be due to vitamin E as the majority of other trials also provided vitamin C. Few trials were conducted in women known to have inadequate intakes of vitamin E, and thus the findings may not be generalizable to all populations.

The evidence reviewed here does not support the routine use of vitamin E in pregnancy for the prevention of fetal or neonatal death, preterm birth, preeclampsia, or intrauterine growth restriction for women with or without a high risk of adverse pregnancy outcomes. While vitamin E supplementation reduced the risk of placental abruption, an increased risk of term PROM and the side effect abdominal pain was also found among women receiving vitamin E.

Further research into the effect of vitamin E on the risk of placental abruption is warranted. Furthermore, additional research is needed into the effects of vitamin E supplementation in women with inadequate intakes of vitamin E. Follow-up of women and children in previous or ongoing trials would provide insight into the potential long-term effects of vitamin E supplementation during pregnancy.