Interventions with Vitamins B6, B12 and C in Pregnancy

Authors


Daphna Dror, Allen Laboratory, USDA, ARS Western Human Nutrition Research Center, 430 W. Health Sciences Dr., Davis, CA 95616, USA. E-mail: dkdror@ucdavis.edu

Abstract

The water-soluble vitamins B6, B12 and C play important roles in maternal health as well as fetal development and physiology during gestation. This systematic review evaluates the risks and benefits of interventions with vitamins B6, B12 and C during pregnancy on maternal, neonatal and child health and nutrition outcomes. Relevant publications were identified by searching PubMed, Popline and Web of Science databases. Meta-analyses were conducted for outcomes where results from at least three controlled trials were available. Potential benefits of vitamin B6 supplementation were reduction in nausea and vomiting, improvement in dental health, and treatment of some cases of anaemia. In meta-analysis based on three small studies, vitamin B6 supplementation had a significant positive effect on birthweight (d = 217 g [95% confidence interval (CI) 130, 304]). Interventions with vitamin C alone or combined with vitamin E did not systematically reduce the incidence of pre-eclampsia, premature rupture of membranes, or other adverse pregnancy outcomes. In meta-analyses, vitamins C and E increased the risk of pregnancy-related hypertension (relative risk 1.10 [95% CI 1.02, 1.19]). Effects of vitamin B6 or C intervention on other neonatal outcomes, including preterm birth, low birthweight, and perinatal morbidity and mortality, were not significant. Data on child health outcomes were lacking. Despite the prevalence of vitamin B12 deficiency amongst populations with limited intake of animal source foods, no intervention trials have evaluated vitamin B12 supplementation before or during pregnancy. In conclusion, existing evidence does not justify vitamin C supplementation during pregnancy. Additional studies are needed to confirm positive effects of vitamin B6 supplementation on infant birthweight and other outcomes. While vitamin B12 supplementation may reduce the incidence of neural tube defects in the offspring based on theoretical considerations, research is needed to support this hypothesis.

Maternal nutritional status and intake during pregnancy are critical to maternal, neonatal and child health. The water-soluble vitamins B6, B12 and C play important roles in maternal health as well as fetal development and physiology, with deficiencies or depletion during pregnancy potentially leading to adverse haematological, neurological and growth consequences. The purpose of this systematic review is to evaluate the efficacy, benefit and safety of interventions with vitamins B6, B12 and C during pregnancy. Specific health outcomes considered include low birthweight due to intrauterine growth restriction, preterm birth, neonatal growth, morbidity and mortality, child growth, morbidity and mortality, maternal morbidity and mortality and maternal nutritional status.

Vitamin B6

Background

Vitamin B6 comprises six related pyridine derivatives: pyridoxal (PL), pyridoxine (PN), pyridoxamine (PM) and their respective 5′-phosphates (PLP, PNP and PMP). The main form of vitamin B6 in animal tissue is PLP, which acts as a coenzyme for numerous enzymes involved in amino acid metabolism; haem, nucleic acid, hormone, neurotransmitter, carbohydrate and lipid biosynthesis; trans-sulfuration of homocysteine to cysteine, and release of glucose from stored glycogen.1 Food sources of vitamin B6 include fortified cereals, potatoes, bananas, meat, fish and poultry, with animal foods containing primarily PLP or PMP and plant foods and supplements containing PN or PNP.1,2 Bioavailability of vitamin B6 is estimated to be >75% from food in a mixed Western diet and >90% from supplements. Classic symptoms of vitamin B6 deficiency are microcytic anaemia, seborrheic dermatitis, epileptiform convulsions, depression and confusion, most of which are related to the role of vitamin B6 as a coenzyme in haemoglobin and neurotransmitter biosynthesis.

Vitamin B6 in pregnancy

Plasma PLP, the best single indicator of vitamin B6 status, declines during pregnancy more than can be accounted for by increase in blood volume, with the most significant drop during the third trimester when blood volume plateaus. The reduction in PLP concentration may represent normal physiological change in pregnancy as there have been no reports of clinical symptoms of deficiency in pregnant women with decreased status indicators.2 In contrast, fetal plasma PLP concentrations in the second and third trimester and at term are significantly higher than in the mother, suggesting fetal sequestration of the vitamin.3,4 Breast milk vitamin B6 concentrations are low in the first days post-partum and subsequently increase for the duration of lactation,5 with a significant positive correlation between maternal plasma PLP and breast milk vitamin B6 demonstrated at 1, 4 and 6 months post-partum.6

Teratogenic effects of vitamin B6 deficiency have not been demonstrated in humans, although in animal studies severe maternal deficiency has produced offspring with lower body weight, skeletal defects, convulsions and impaired neuromotor development.7,8 In humans, vitamin B6 deficiency leads to hyperhomocysteinaemia, which has been related to obstetric complications such as pre-eclampsia.9 It has been hypothesised that vitamin B6 deficiency may contribute to aggravation of gingivitis and oral tissue changes associated with pregnancy.10,11

Vitamin B6 deficiency may account for some types of anaemia in pregnancy, as demonstrated by a study of 56 anaemic Japanese pregnant women who were non-responsive to iron supplementation but responded to vitamin B6 therapy.12 A prospective observational study in China found a decreased probability of conception and an increased risk of early pregnancy loss in women in the lowest quartile of vitamin B6 status and in women with vitamin B6 deficiency.13

Because plasma PLP decreases with age and gestation,14 vitamin B6 deficiency in pregnancy is not clearly defined and there are few estimates of its prevalence. In the US, 6–17% of women taking multivitamins containing 10 mg/day vitamin B6 were deficient at various stages of pregnancy,15 while 26% of unsupplemented Chinese women planning pregnancy were vitamin B6 deficient.16 Risk factors for low plasma PLP amongst women of reproductive age include multiple gestation12 and oral contraceptive use.17 The recommended dietary allowance (RDA) for vitamin B6 in pregnancy is 1.9 mg/day, an increase of 0.6 mg/day over the non-pregnant requirement.

Despite isolated case reports of congenital defects in newborns whose mothers received high-dose pyridoxine early in pregnancy,18–20 no evidence of teratogenicity upon supplementation has been shown in controlled animal studies.21–23 Vitamin B6 in combination with the antihistamine doxylamine was prescribed worldwide to treat nausea and vomiting of pregnancy (NVP) in the 1960s and 1970s and is currently available for use in Canada.24 A recent Cochrane review concluded that there is a lack of consistent evidence that this therapy is effective in reduction of symptoms in early pregnancy.25 Doses of 50–510 mg/day taken during the first trimester have not been associated with adverse fetal outcomes.24,26,27 In addition, pyridoxine supplementation has been used therapeutically to suppress prolactin secretion and inhibit lactation,28 though its efficacy has not been demonstrated consistently.29–32

Vitamin B12

Background

Vitamin B12, or cobalamin, comprises a group of cobalt-containing corrinoids that can be converted to methylcobalamin, a coenzyme for methionine synthase, or 5-deoxyadenosylcobalamin, a coenzyme for L-methylmalonyl-CoA mutase. Because of its role in DNA synthesis, cobalamin is essential for cell multiplication during pregnancy.33 In the diet, naturally occurring protein-bound vitamin B12 is found exclusively in animal products such as meat, fish, eggs and milk while synthetic crystalline B12 is used as a fortificant in cereals.2 In healthy adults with normal gastric function, fractional absorption of protein-bound vitamin B12 is 50% of a 1 µg dose and decreases to 5% of a 25 µg dose.34 Vitamin B12 intakes and serum concentrations reflect dietary habits and increase progressively from vegans who consume no animal source foods to lacto-ovo vegetarians, vegetarians who eat fish, and omnivores who consume meat.35 The primary causes of vitamin B12 deficiency are low intake and malabsorptive disorders, which may be due to pernicious anaemia, achlorhydria, ileal damage or gastric bypass.36 Symptoms of vitamin B12 deficiency are haematological, neurological and cognitive, including megaloblastic anaemia, tingling and numbness of the extremities, gait abnormalities, visual disturbances, memory loss and dementia. Because of reabsorption from bile during enterohepatic circulation, adults starting with adequate stores and normal absorption may tolerate a vitamin B12 deficient diet or malabsorptive disorder for years before developing clinical symptoms of deficiency. Furthermore, folic acid fortification or supplementation with >1 mg/day may ‘mask’ clinical symptoms of vitamin B12 deficiency such as anaemia by exogenously replacing the folate trapped in the form of 5-methyl-tetrahydrofolate, unable to be demethylated in the absence of B12-dependent methionine synthase.1

Vitamin B12 in pregnancy

Serum vitamin B12 concentrations decline during pregnancy more than can be accounted for by haemodilution.37 There is some evidence of increased B12 absorption during pregnancy, with newly absorbed vitamin B12 being more important to placental transport than maternal liver stores.38 Vitamin B12 is concentrated in the placenta and transferred to the fetus down a concentration gradient, with newborn vitamin B12 concentrations approximately double those of the mother.38 Vitamin B12 concentrations are lower in the breast milk of mothers with deficient or marginal serum vitamin B12 status than in mothers with adequate status, though results from several studies suggest that maternal intake is a stronger predictor of breast milk vitamin B12 concentration than maternal status.39–42

The prevalence of vitamin B12 deficiency in pregnancy is difficult to quantify given the physiological decline in serum vitamin B12 concentration throughout gestation. Prevalence estimates have varied from 5% in Canadian women <28 days gestation to 72% in Turkish women immediately prior to delivery.43,44 Several studies have found that despite depressed cobalamin concentrations, homocysteine levels were not elevated as would be expected in vitamin B12 deficiency.37,45,46 A concomitant increase in erythrocyte cobalamin and decrease in saturation of cobalamin binding serum proteins suggest a redistribution of cobalamin during pregnancy.46 Nevertheless, because a number of studies have revealed that vitamin B12 deficiency is prevalent across population sectors,43,44,47–50 it is likely that many women worldwide begin pregnancy with inadequate B12 stores which are not further repleted.

Vitamin B12 deficiency during pregnancy has been associated with adverse fetal and neonatal outcomes, most notably neural tube defects (NTDs)51–54 and delayed myelination or demyelination.55–57 In addition, a prospective study in India found a significant association between low maternal vitamin B12 and high folate status during pregnancy and adiposity and insulin resistance in 6-year-old offspring.58

During pregnancy the total vitamin B12 requirement of the fetus is estimated to be 50 µg while maternal stores in women with a mixed diet are estimated at >1000 µg.59 In a well-nourished woman, therefore, body stores of B12 are adequate to meet fetal needs during gestation. In many developing regions of the world, however, low intake of animal source foods compromises vitamin B12 status in women of reproductive age.60 The RDA for vitamin B12 in pregnancy is 2.6 µg/day, 0.2 µg/day greater than the corresponding RDA for non-pregnant women and adolescents.1

No adverse effects on pregnancy outcomes have been associated with a high vitamin B12 intake from food or supplements in healthy individuals. Therapeutic doses of vitamin B12 orally (500–1000 µg/day) or parenterally (>1 mg dose monthly or every few months) are used in standard clinical practice to treat patients with pernicious anaemia. Only a small fraction (≤5%) of orally administered high-dose vitamin B12 can be absorbed from the gastrointestinal tract.34 Because of insufficient evidence of adverse outcomes, no tolerable upper limit for vitamin B12 has been set for the general population or for pregnant women.

Vitamin C

Background

Vitamin C, or ascorbic acid, is an electron donor for eight human enzymes involved in biosynthesis of collagen, carnitine, hormones and neurotransmitters and acts as an antioxidant in quenching reactive oxygen and nitrogen species in an aqueous environment.61 The main dietary sources of vitamin C are fruits and vegetables, particularly citrus fruits, strawberries, tomatoes and potatoes. The vitamin C content of fruits and vegetables varies depending on growing conditions, season of the year, stage of maturity, location, storage time and cooking practices.62 Approximately 70–90% of typical intake from food or supplements (30–180 mg/day) is absorbed,63–65 predominantly through a sodium-dependent active transport process.66 At intakes above 1 g/day, absorption efficiency decreases to ≤50% and occurs mainly by simple diffusion.63,66 The classical manifestation of severe vitamin C deficiency is scurvy, which is characterised by inflamed and bleeding gums, petechiae, ecchymoses, perifollicular haemorrhages, weakness, fatigue and depression.67–69 Marginal vitamin C status has been associated with increased risk of cardiovascular disease and decreased immune function.70–72

Vitamin C in pregnancy

Plasma vitamin C concentrations fall throughout pregnancy, likely due to haemodilution as well as active transfer to the fetus.73 The placenta clears the oxidised form of the vitamin from maternal circulation and delivers the reduced form to the fetus.74

In pregnant women, vitamin C deficiency has been implicated in increased risk of infections, premature rupture of membranes (PROM), preterm birth, pre-eclampsia and eclampsia.75–80 However, many studies demonstrating such associations have been cross-sectional, precluding inference about causality. Based on experimental results in animals, it has been suggested that gestational and neonatal vitamin C deficiency may impair normal brain development since the brain is particularly susceptible to oxidative damage.81 However, a prospective study of antioxidant vitamin concentrations in maternal and cord blood at delivery found no correlations between vitamin C status and intellectual development evaluated at 2 years of age.82

The prevalence of vitamin C deficiency in late pregnancy has been estimated to be 30.8% in Brazil83 and 61–69% in a multi-centre Chinese study.84 In the general population, recent estimates of prevalence of deficiency (serum or plasma ascorbic acid <11 µmol) in women of childbearing age are 16% in a representative sample of low-income women in the UK from 2003 to 2005,85 10.4% in low-income and 5.0% in high-income women in the US 2003 to 2004 NHANES survey,86 and 12.6% in non-smokers aged 20–29 in Toronto from 2004 to 2008.87 Risk factors for deficiency include low dietary intake, oral contraceptive use, obesity and smoking.86–88

The total amount of vitamin C required for fetal development is unknown. Based on evidence that 7 mg/day prevents infants from developing scurvy,89–91 the RDA calculated to meet maternal and fetal needs is 85 mg/day for adult women, 5 mg/day greater than the corresponding RDA for non-pregnant women.61 The vitamin C content of breast milk is correlated with maternal intake and status and varies by season.

There is no evidence of maternal toxicity or teratogenic effects of excess vitamin C intake during pregnancy.61 However, because vitamin C is actively transported to the fetus, there is potential for excess maternal intake to lead to elevated fetal plasma concentrations.

Methods

To identify original research articles describing interventions with vitamins B6, B12 and C during pregnancy, the PubMed database was used to search for relevant Medical Subject Heading (MeSH) and general terms. MeSH search terms included for each of the vitamins were as follows: vitamin B6: ‘pyridoxal’, ‘pyridoxamine’, ‘pyridoxine’, ‘vitamin B6’; vitamin B12: ‘vitamin B12’; and vitamin C: ‘ascorbic acid’. In addition, the general terms ‘cobalamin’ and ‘vitamin C’ were included in searches for vitamin B12 and vitamin C respectively. Included for all vitamins were the MeSH terms or phrases ‘pregnancy’, ‘birthweight’, ‘growth’ and ‘fetal growth retardation’, and the general terms ‘supplementation’, ‘preterm’ and ‘intrauterine’. The same terms were used to conduct searches using other population and health databases Popline and Web of Science.

All identified studies pertaining to supplementation with vitamins B6, B12 or C during pregnancy with outcomes including maternal, infant or child nutritional status, growth, morbidity, or mortality were reviewed for target population and concomitant interventions or exposures. Studies conducted in non-pregnant populations or not reporting outcomes of interest were excluded, while observational cohort, quasi-randomised and uncontrolled trials meeting subject and methodological criteria were included. An additional search was used to find review articles, which were examined for references to other relevant studies.

A standardised abstraction table (available from the corresponding author upon request) was used to extract and organise key information from experimental and observation studies that met the selection criteria. As part of this protocol, study level variables including design, setting, target population and limitations were summarised and each outcome was considered separately in terms of definition, sample size, and crude and adjusted results. Study quality was assessed using a modified version of Grading of Recommendations Assessment, Development, and Evaluation (GRADE)92,93 (Tables 1 and 2; Tables S1–3).

Table 1.  Summary of findings and overall assessment of quality of evidence, by outcome, for interventions with vitamin B6
Quality assessmentSummary of findings
 Directness
No. of studiesDesignLimitationsConsistencyGeneralisability to population of interestGeneralisability to intervention of interestNo. participants at risk (exposed group)Effect estimate
  1. NS, not significant; PN-HCl, pyridoxine hydrochloride; QE, quasi-experimental; RCT, randomised controlled trial; RR, relative risk; (+), positive; (−), negative.

Birthweight (g): overall quality of evidence = low
52 RCTs, 1 QE, 1 prospective and 1 retrospective cohortRCTs with small sample size, one with large loss-to-follow-up; no placebo, randomisation in QE; observational studies with B6 + doxylaminePoor; RCT found (−) effect, 1 cohort found (+) effect but not after adjusting for maternal weight, other studies found no effectAll studies conducted in developed countriesRCTs and QE involved supplementation with 2.6–25 mg/day PN-HCl; observational studies involved 10–510 mg/day alone or in conjunction with doxylamine and other NVP therapies130 (3 studies reporting results of intervention vs. control group, of which 1 included doxylamine)Difference and 95% CI (g): 217 [130, 304]
Toxaemia (pre-eclampsia and eclampsia): overall quantity of evidence = very low
21 RCT and 1 QEQE controls not concurrent, no placebo. In RCT intervention changed during studyQE found significant protective effect of treatment; RCT found no effectBoth studies conducted in developed countriesStudies involved supplementation with 10–20 mg PN-HCl1366RR and 95% CI: 0.81 [0.00, 67.04]
Fetal malformations: overall quality of evidence = low
21 prospective cohort and 1 case controlBoth studies with possible selection and recall biasNeither study found significant treatment effectBoth studies conducted in developed countriesSupplementation with 10–510 mg/day PN-HCl in prospective cohort, ‘pyridoxine use’ in case control2109RR and 95% CI: 0.82 [0.28, 2.42]
Birth length: overall quality of evidence = very low
21 RCT and 1 QERCT with large loss-to-follow-up and small sample size. QE failed to report numeric resultsNeither study found significant treatment effectBoth studies conducted in developed countries1–20 mg/day PN-HCl16 (only RCT reported numeric results)NS
Preterm birth: overall quality of evidence = very low
11 QEControls not concurrent, no placeboNo significant treatment effectStudy conducted in developed country10 mg/day PN-HCl410NS
Stillbirth: overall quality of evidence = very low
11 QEControls not concurrent, no placeboNo significant treatment effectStudy conducted in developed country10 mg/day PN-HCl410NS
Delayed/missing/filled teeth rating: overall quality of evidence = very low
11 RCTMethods poorly describedSignificant (+) treatment effect of lozengesStudy conducted in developed country20 mg/day PN-HCl in pill or lozenge form367Difference 0.53 units, P = 0.01
Neurodevelopment: overall quality of evidence = very low
11 prospective cohortResults cannot be attributed solely to B6, small sample sizeSignificant (+) treatment effectStudy conducted in developed countryTreatment of NVP with doxylamine-B6; variable dose, duration and frequency92Neurocogintive test scores significantly higher in treatment group, p < 0.02
Table 2.  Summary of findings and overall assessment of quality of evidence, by outcome, for interventions with vitamins C and E
Quality assessmentSummary of findings
No. of studies and study designHeterogeneity of results?Consistent size of effect?Generalisable to pop of interest?Generalisable to intervention of interest?Other sources of bias (e.g. major limitations in study design)No. participantsStatistical methodEffect estimate
  1. Fav, favourable; N/A, not applicable; part, participants; Pop, population; preg, pregnancy; RCT, randomised controlled trial; sig, significant(ly); supp, supplemented; Unfav, unfavourable.

Pre-eclampsia: overall quality of evidence = high
9 RCTsLow (I2 = 14%); 1 study showed sig fav effect of interventionOnly 1 study showed an effect4 of 9 studies in low-income pop, 3 in middle/low-income countriesAll studies supp with 1000 mg C + 400 IU E/day6 of 9 studies in part at increased risk for pre-eclampsia19 798Risk ratio (random, 95% CI)1.00 [0.91, 1.10]
Premature rupture of membranes: overall quality of evidence = moderate
4 RCTsYes (I2 = 83%); 1 sig fav, 2 sig unfav, 1 with no sig effect of interventionNo, effects in opposite directions3 of 4 studies in pop in middle/low-income countriesAll studies supp with 1000 mg C + 400 IU E/day1 of 4 studies in part at increased risk for preg complications13 026Risk ratio (random, 95% CI)1.11 [0.39, 3.19]
Pre-term premature rupture of membranes: overall quality of evidence = moderate
4 RCTsYes (I2 = 54%); 1 study showed a sig unfav effect of interventionOnly 1 study showed an effect1 study in low-income pop in middle-income countryAll studies supp with 1000 mg C + 400 IU E/day2 of 4 studies in part at increased risk for preg complications13 293Risk ratio (random, 95% CI)1.10 [0.62, 1.96]
Pregnancy-related hypertension: overall quality of evidence = moderate
5 RCTsNo (I2 = 0%), 2 studies with sig unfav effect of interventionOnly 2 studies showed an effect1 study in pop in middle-income countryAll studies supp with 1000 mg C + 400 IU E/day2 of 5 studies in part at increased risk for preg complications17 353Risk ratio (random, 95% CI)1.10 [1.02, 1.19]
Preterm birth (<37 wk): overall quality of evidence = high
9 RCTsYes (I2 = 51%); 1 sig fav, 1 sig unfav, others no effect of intervention2 studies with effect, opposite directions4 of 9 studies in low-income pop, 3 in middle/low income countries7 supp with 1000 mg C + 400 IU E/day, 1 with 500 mg/day C and 1 with 100 mg/day C1 trial aborted early. 5 of 9 studies in part at increased risk for preg complications19 632Risk ratio (random, 95% CI)1.00 [0.89, 1.12]
Birthweight < 2500 g: overall quality of evidence = moderate
5 RCTsYes (I2 = 52%), no studies with sig effect of interventionN/A3 of 5 studies in low-income pop, 2 in middle/low income countriesAll studies supp with 1000 mg C + 400 IU E/day4 of 5 studies in part at increased risk for preg complications14 878Risk ratio (random, 95% CI)0.99 [0.83, 1.19]
Birthweight < 1500 g: overall quality of evidence = low
3 RCTsLow (I2 = 19%); 1 study showed sig unfav effect of interventionOnly 1 study showed an effect2 of 3 studies in low-income pop in middle/low income countriesAll studies supp with 1000 mg C + 400 IU E/dayAll studies conducted in part at increased risk for preg complications4997Risk ratio (random, 95% CI)1.01 [0.58, 1.73]
Small-for-gestational age (<10th percentile): overall quality of evidence = high
8 RCTsYes (I2 = 35%), no studies showed sig effect of interventionN/A4 of 8 studies in low-income pop, 3 in middle/low income countriesAll studies supp with 1000 mg C + 400 IU E/day6 of 8 studies conducted in part at increased risk for preg complications10 164Risk ratio (random, 95% CI)0.94 [0.82, 1.09]
Stillbirth: overall quality of evidence = moderate
7 RCTSNo (I2 = 0%), no studies showed sig effect of interventionN/A4 of 7 studies in low-income pop, 3 in middle/low income countries6 studies supp with 1000 mg C + 400 IU E/day, 1 with 500 mg/day C1 trial aborted early. 5 of 7 studies in part at increased risk for preg complications8912Risk ratio (random, 95% CI)1.23 [0.85, 1.94]
Neonatal death: overall quality of evidence = high
9 RCTsNo (I2 = 0%), no studies showed sig effect of interventionN/A5 of 9 studies in low-income pop in middle/low income countries8 studies supp with 1000 mg C + 400 IU E/day, 1 with 500 mg/day C1 trial aborted early. 6 of 9 studies in part at increased risk for preg complications20 053Risk ratio (random, 95% CI)0.84 [0.63, 1.13]
Birthweight (g): overall quality of evidence = high
8 RCTsYes (I2 = 93%), no studies showed sig effect of interventionN/A3 of 8 studies in low-income pop in middle/low income countries6 supp with 1000 mg C + 400 IU E/day, 1 with 500 mg/day C and 1 with 100 mg/day C1 trial aborted early. 5 of 8 studies in part at increased risk for preg complications16 365Mean difference (IV, fixed, 95% CI)−26 [−82, 29]

Meta-analyses were planned for outcomes where at least three controlled trials were available with sufficient information to estimate effect sizes. Random effects models were used to evaluate binary and continuous outcomes in the total data sets.

Results

Interventions with vitamin B6 during pregnancy

Vitamin B6 is often included in multivitamin supplements administered routinely or experimentally during pregnancy. However, only a limited body of research has focused on the effects of exclusive vitamin B6 supplementation during pregnancy on maternal and infant outcomes. In total, data were abstracted from 11 observational, case–control and intervention studies of vitamin B6 (Tables 3a and 3b).

Table 3a.  Randomised controlled trials with vitamin B6 Thumbnail image of
Table 3b.  Other interventions with vitamin B6 Thumbnail image of

Maternal outcomes

Maternal outcomes that have been investigated in association with vitamin B6 supplementation include risk of pre-eclampsia and dental decay, reduction in breast milk production, and changes in biomarkers of vitamin B6 status. One US quasi-experimental trial found a significant reduction in incidence of pre-eclampsia in women treated with 10 mg/day pyridoxine-hydrochloride compared with non-placebo controls.94 However, another placebo-controlled randomised trial (RCT) found no significant difference in the incidence of pre-eclampsia (toxaemia) in women treated with 20 mg/day pyridoxine.95 A single RCT investigating the effect of vitamin B6 supplementation during pregnancy on decayed, missing or filled teeth (DMF) rating found that participants receiving 20 mg/day vitamin B6 had a significantly smaller increase in DMF rating than women receiving placebo.96 Supplementation with 2–6 mg/day pyridoxol during pregnancy did not have an antilactogenic effect in a retrospective cohort of 11 Swedish mothers,31 nor did treatment of 14 women in Mexico with 450 mg/day pyridoxine during the first week post-partum inhibit lactation.29

Maternal plasma PLP increased in women supplemented with vitamin B6 during pregnancy across trials reporting this outcome.95,97–99 Several studies have measured stimulation of erythrocyte aminotransferases, another indicator of vitamin B6 status. While one RCT found that supplementation with vitamin B6 had a significant effect on stimulation of aspartate aminotransferase,97 another prospective cohort study failed to find a correlation between maternal plasma PLP and aspartate or alanine aminotransferase at various levels of supplementation (2.5–10 mg/day).98 Absolute erythrocyte aminotransferase activity was not significantly different between supplementation and control groups in one RCT,97 though absolute enzyme activities vary widely and have limited use as indicators of status.1

Neonatal outcomes

Neonatal outcomes that have been evaluated in trials supplementing pregnant women with vitamin B6 alone include incidence of birth defects, birthweight, risk of preterm birth, risk of stillbirth and cord blood vitamin B6 status. A large-scale case–control study of 25 congenital abnormalities and pyridoxine use in early pregnancy in Hungary found no evidence of teratogenic risk of supplementation but rather a protective effect of pyridoxine against cardiovascular defects, undescended testis and clubfoot.100 In another case–control study in the Netherlands, periconceptional intake of pyridoxine based on a food frequency questionnaire was significantly lower in mothers of offspring with orofacial cleft.101

In a quasi-experimental study, there was no difference in incidence of stillbirth between vitamin B6 supplemented and control groups.94 Several studies have found increases in birthweight24,99,102 and serum or plasma PLP in cord blood97–99 following maternal supplementation.

Child outcomes

Data on morbidity, mortality, nutritional status and growth of children whose mothers received vitamin B6 supplements during pregnancy are not available. In a single study following offspring of Canadian mothers supplemented in pregnancy, children age 3–7 years born to mothers treated with vitamin B6–doxylamine scored significantly higher on several neurocognitive tests and had fewer sleep problems than children of mothers with nausea and vomiting who had not received therapy.104

Meta-analysis of vitamin B6 interventions during pregnancy

Amongst vitamin B6 interventions, birthweight was the only outcome warranting meta-analysis with results from three controlled studies.24,97,102 The effect of vitamin B6 supplementation on birthweight was significant (difference and 95% confidence interval: 217 g [130, 304]), but this result should be interpreted with caution as sample size was small (total n = 247) and one study included supplementation with the antihistamine doxylamine as well as confounding by maternal weight (Table 4).

Table 4.  Meta-analysis of vitamin B6 interventions during pregnancy (studies included and results)
OutcomeReference (first author and year)nMean (SD) in gP-valueOverall difference [95% CI]Overall P-value
Intervention/control
BirthweightSwartwout, 1960102333130(280)/2893(280)0.02217 g [130, 304]0.009
Schuster, 198497223287(429)/3240(505)NS
Shrim, 2006241923542(512)/3321(512)0.01

Interventions with vitamin B12 during pregnancy

Vitamin B12 as an exclusive supplement has not been investigated during pregnancy. Many multivitamin supplementation trials have included vitamin B12; however, any effects on maternal, infant and childhood outcomes cannot be attributed to any specific vitamin.

Several studies have found a relationship of maternal vitamin B12 status with intrauterine growth retardation and birthweight.105,106 However, neither serum cobalamin nor homocysteine were significant determinants of birthweight in a prospective study measuring maternal concentrations at 30–34 weeks' gestation.107

Interventions with vitamin C during pregnancy

Most interventions with vitamin C during pregnancy have included supplementation with vitamin E since both vitamins are potent physiological antioxidants. Nevertheless, treatment effects cannot be attributed directly to vitamin C. In total, data were abstracted from two intervention trials with vitamin C exclusively and 13 interventions trials with a combination of vitamin C and vitamin E during pregnancy (Table 5).

Table 5.  Interventions with vitamin C or vitamin C and vitamin E
Reference (first author and year)LocationStudy designStudy populationSubject sample sizeInterventionOutcomes measuredResultsLimitations
  1. BW, birthweight; CM, congenital malformations; cont, control(s); del, delivery; diff, difference; DM, diabetes mellitus; gest, gestation(al); HTN, hypertension; hx, history; LBW, low birthweight (<2500 g); mat, maternal; N/A, not applicable; OR, odds ratio; preg, pregnant; PPROM, preterm premature rupture of chorioamniotic membranes; PROM, premature rupture of chorioamniotic membranes; RCT, randomised controlled trial; RR, relative risk; SES, socio-economic status; SGA, small-for-gestational age (<10th percentile); sig, significant(ly); supp, supplementation; vit, vitamin; VLBW, very low birthweight (<1500 g); wk, weeks.

Gulmezoglu, 1997120Johannesburg, South AfricaRCTPreg women hospitalised with early onset (24–32 wk) severe pre-eclampsia56: 27 int/29 placebo1000 mg vit C, 800 IU vit E and 200 mg allopurinol/day through delStillbirth, neonatal death and BWNo sig diff in stillbirth, neonatal death or BW between int and cont groupsGest age at del not reported, small sample size, late treatment initiation, hospitalised population not generalisable
Chappell, 1999108London, UKRCTPreg women 16–22 wk gest at high risk of pre-eclampsia (Doppler or hx)283: 141 int/142 placebo1000 mg vit C and 400 IU vit E/day through delPre-eclampsia, BW, SGASig lower odds of pre-eclampsia in vit group. No sig diff in BW or SGACompliance not monitored, adjustment of OR not explained, participants at high risk for pre-eclampsia
Steyn, 2003121Tygerberg, South AfricaRCTPreg women <26 wk gest with hx of late abortion or preterm labour200: 100 int/100 cont500 mg vit C/day until 34 wk gestPreterm birth, perinatal death, BWPreterm del risk sig higher in vit group. No sig diff in perinatal death or BWShort supp dose and duration, compliance not monitored, high-risk preg
Pressman, 2003119Rochester, NY, USARCTPreg women scheduled to undergo planned Cesarean del19: 10 int/9 cont500 mg vit C and 400 IU vit E/day from 35 wk gest through delMat and cord plasma and amniotic fluid vit CNo sig diff in mat or cord plasma or amniotic fluid vit C concentrationsCompliance not reported, short supp duration, 8–12 h delay between last dose and del so possible excretion
Casanueva, 200573Mexico City, MexicoRCTPreg women <20 wk gest with no acute or chronic disease109: 52 int/57 cont100 mg vit C/day through delPROM, mat plasma and leucocyte vit C, preterm del, BWSig lower risk of PROM in vit group. Plasma and leucocyte vit C decreased sig in both groups. No sig diff in preterm del or BWCompliance not reported
Beazley, 2005109US (specific location N/A)RCTPreg women 14–20 wk gest with hx of pre-eclampsia, HTN, DM or multiple gest100: 52 int/48 cont1000 mg vit C and 400 IU vit E/day through delPre-eclampsia, preterm del, LBW, SGANo sig diff in any outcomeDesired sample size not achieved, compliance not reported, participants at high risk of pre-eclampsia
Poston, 2006110UK (national)RCTPreg women 14–22 wk gest at high risk for pre-eclampsia2395: 1196 int/1199 placebo (Offspring: 1393/1391)1000 mg vit C and 400 IU vit E/day through delPre-eclampsia, gest HTN, LBW, VLBW, SGA, preterm del, perinatal death, BW, mat plasma CVit group had sig higher risk of gest HTN, requiring IV anti-HTN therapy and LBW. Mat plasma C sig higher in vit group at 20–36 wk. Other outcomes not sig diffParticipants at increased risk of pre-eclampsia (incidence 26% vs. estimated 2% in society)
Rumbold, 2006111Adelaide, AustraliaRCTNulliparous preg women 14–22 wk gest with singleton fetus and normal blood pressure1877: 935 int/942 placebo1000 mg vit C and 400 IU vit E/day through delPre-eclampsia, gest HTN, birth length, BW, PPROM, SGA, preterm del, stillbirth, neonatal deathNo sig diff for any outcome. Women in vit group more likely to be hospitalised for anti-HTN therapy (RR 1.54 [1.00, 2.39])Majority of women with baseline vit C and E intakes above RDA, so results not generalizable to populations with poorer intake
Spinnato, 20071124 cities, BrazilRCTPreg women 12–20 wk gest with chronic HTN or pre-eclampsia hx707: 355 int/352 placebo1000 mg vit C and 400 IU vit E/day through delPre-eclampsiaNo sig diff in risk of pre-eclampsiaStudy was underpowered, participants at high pre-eclampsia risk
Spinnato, 20081174 cities, BrazilRCTPreg women 12–20 wk gest with chronic HTN or pre-eclampsia hx697: 349 int/348 placebo1000 mg vit C and 400 IU vit E/day through delPROM, SGA, PPROM, BW, LBW, VLBW, stillbirth, neonatal deathSig greater risk of PROM and PPROM in vit group. No sig diff for other outcomesStudy was underpowered, participants at high pre-eclampsia risk
Villar, 2009113India, Peru, South Africa and VietnamRCTPreg women 14–22 wk gest of low SES and high risk for pre-eclampsia1355: 681 int/674 placebo1000 mg vit C and 400 IU vit E/day through delPre-eclampsia, LBW, VLBW, SGA, preterm del, perinatal death, CMNo sig diff for any outcomeParticipants at high pre-eclampsia risk, low SES and low nutritional status based on community survey
Roberts, 2010114US (national)RCTNulliparous preg women 9–16 wk gest with no pre-eclampsia risk factors9969: 4993 int/4976 placebo1000 mg vit C and 400 IU vit E/day through delPre-eclampsia, gest HTN, LBW, PROM, SGA, preterm del, BW, neonatal deathSig greater risk of gest HTN in vit group. No sig diff for other outcomesNearly 80% of women already taking prenatal vit that contained average 100 mg vit C and 22 IU vit E/day
Hauth, 2010118US (national)RCTNulliparous preg women 9–16 wk gest (same participants as Roberts, 2010)9968: 4992 int/4976 placebo1000 mg vit C and 400 IU vit E/day through delPPROM, preterm del, preterm del attributable to PPROMSig lower odds of PPROM preterm birth attributable to PPROM at <32 wk (but not <35 or <37 wk) in vit groupSig result at <32 wk may be attributable to clinical imprecision in differentiating PPROM preterm labour
McCance, 2010116Ireland, UKRCTPreg women 8–22 wk gest with type I DM749: 375 int/374 placebo1000 mg vit C and 400 IU vit E/day through delPre-eclampsia, gest HTN, SGA, LBW, PPROM, preterm del, stillbirth, neonatal death, CM, BWSig lower risk of preterm del in vit group. In subgroup of women with baseline ascorbate <10 µmol/L, sig lower risk of pre-eclampsia in vit group. No other sig diffN/A
Xu, 2010115Canada, MexicoRCTPreg women 12–18 wk gest2363: 1167 int/1196 placebo1000 mg vit C and 400 IU vit E/day through delPre-eclampsia, gest HTN, SGA, PROM, PPROM, preterm del, stillbirth, neonatal deathSig higher risk of PROM and PPROM in vit group. No sig diff for other outcomesN/A

Maternal outcomes

Maternal outcomes that have been investigated in association with vitamin C and E supplementation include pregnancy complications (gestational hypertension, pre-eclampsia and PROM) and maternal vitamin C status. Despite an initial British RCT suggesting that vitamin C and E supplementation was effective at reducing the risk of pre-eclampsia in 283 women at high risk based on abnormal Doppler exam or history of the disorder,108 other larger-scale RCTs conducted amongst women at low or high risk of pre-eclampsia in developed and developing countries have failed to corroborate this result.109–115 Several of these studies have found a significantly higher risk of gestational hypertension requiring antihypertensive therapy in the vitamin group.110,111,114 In the UK, women with type I diabetes preceding pregnancy randomly allocated to receive vitamin C and E or placebo showed no difference in rates of pre-eclampsia overall, though the risk of pre-eclampsia was significantly higher amongst the subset of women with baseline plasma ascorbate concentration <10 µmol/L.116

While one small RCT found a significantly lower relative risk of PROM in the group supplemented with vitamin C,73 two multi-centre RCTs found that daily supplementation with vitamins C and E increased the risk of preterm and term PROM114,116 and two similar RCTs found no difference in risk of preterm PROM between groups.110,115 In a large multi-centre RCT involving 10 154 nulliparous pregnant women, daily supplementation with vitamins C and E resulted in less frequent preterm PROM before 32 weeks' gestation but not before 35 or 37 weeks' gestation. The authors conjectured that the single significant finding may have been an artefact of clinical imprecision in estimating the timing of membrane rupture in relation to onset of labour.117

Several RCTs found higher leucocyte73 and plasma vitamin C concentrations in women supplemented with vitamin C and E compared with control women at multiple time points during pregnancy,110,116 while some have failed to find a significant difference in plasma vitamin C concentrations between groups during pregnancy73 or in plasma or amniotic fluid vitamin C concentrations at delivery.119

Neonatal outcomes

No significant difference was found in neonatal outcomes of women supplemented with vitamin C alone or in combination with other supplements compared with a placebo on the risk of stillbirth, perinatal or neonatal death,108,110,111,113–117,120,121 though one RCT found a significantly increased risk of fetal loss, a composite of spontaneous abortion, stillbirth or neonatal death, in the supplemented group.115 No significant difference was found in the risk of congenital malformations between treatment and control groups.113,116 Most RCTs have found no significant difference in risk of preterm birth (delivery prior to 37 weeks' gestation) between vitamin supplemented and control groups.73,108–111,113–115,117,122 One RCT found an increased risk of preterm delivery in high-risk mothers supplemented with 500 mg/day vitamin C,121 while another found a decreased risk of preterm delivery in mothers with type I diabetes supplemented daily with 1000 mg vitamin C and 400 IU vitamin E.116

Results of most RCTs indicate no effect of vitamin C supplementation during pregnancy on intrauterine growth restriction,108–111,113–117 low birthweight defined as <2500 g,109,114,122 or very low birthweight defined as <1500 g.113,117 One RCT found an increased rate of low birthweight infants in the supplementation group,109 while another showed a higher percentage of infants born small-for-gestational age (<10th percentile) in the placebo group, though statistical significance was not achieved (P = 0.12).108 Birthweight and birth length did not differ significantly between supplementation and control groups in any RCT reporting these outcomes.73,108–111,113,114,116,117,120–122

Child outcomes

Data on morbidity, mortality and growth of children whose mothers received vitamin C supplements during pregnancy are not available. One prospective cohort study of oxidative stress and antioxidant vitamin status in South Korea found that mothers in the highest quartile of vitamin C concentration at 24–28 weeks' gestation had infants with significantly greater weight and length at birth, 6 and 12 months after adjustment for confounders.123

Meta-analyses of vitamin C and E interventions during pregnancy

In a meta-analysis of 14 RCTs with vitamin C, 12 of which included supplementation with vitamin E, no effect of supplementation was found on risk of pre-eclampsia (n = 9), PROM (n = 4), preterm PROM (n = 4), preterm delivery (<37 weeks' gestation, n = 9), low birthweight (<2500 g, n = 5), very low birthweight (<1500 g, n = 3), small-for-gestational age (<10th percentile, n = 8), stillbirth (n = 7) or neonatal death (n = 9). There was a significant treatment effect for gestational hypertension, with more women in the intervention than control groups developing pregnancy-related hypertension (n = 5, P = 0.03). There was no significant difference in mean birthweight by treatment group, nor was birthweight related to risk status for pre-eclampsia. A meta-regression by pre-eclampsia risk status demonstrated similar results, except that the incidence of gestational hypertension was not significantly increased in women at high risk for pre-eclampsia supplemented with vitamins C and E (n = 2, one with women who had type I diabetes prior to pregnancy) (Table 6).

Table 6.  Meta-analysis of vitamins C and E interventions during pregnancy (studies included and results)
OutcomeReference (first author and year)nRR [95% CI]P-valueOverall and high-risk RR [95% CI]Overall and high-risk P-value
Pre-eclampsiaaChappell, 19991082830.39 [0.17, 0.90]b0.021.00 [0.91, 1.10]/0.94 [0.81, 1.09]0.96/0.14
aBeazley, 20051091000.92 [0.40, 2.13]NS
aPoston, 200611023950.97 [0.80, 1.17]NS
Rumbold, 200611118771.20 [0.82, 1.75]NS
aSpinnato, 20071127070.87 [0.61, 1.25]cNS
aVillar, 200911313551.03 [0.85, 1.35]NS
Roberts, 201011499691.07 [0.93, 1.24]NS
aMcCance, 20101167490.81 [0.59, 1.12]NS
Xu, 201011523631.04 [0.75, 1.44]NS
PROMCasanueva, 2005731090.26 [0.08, 0.84]0.021.11 [0.39, 3.19]0.78
aSpinnato, 20081176971.89 [1.11, 3.23]c0.02c
Roberts, 201011498570.96 [0.75, 1.22]NS
Xu, 201011523631.65 [1.23, 2.22]<0.001d
Preterm PROMRumbold, 200611118791.31 [0.77, 2.25]NS1.10 [0.62, 1.96]/1.32 [0.00, 4312.8]0.64/0.88
aSpinnato, 20081176972.68 [1.07, 6.71]c0.03c
Hauth, 201011899680.98 [0.77, 1.25]NS
aMcCance, 20101167490.74 [0.44, 1.24]NS
Pregnancy-related hypertensionaPoston, 200611023951.53 [1.10, 2.13]0.021.10 [1.02, 1.19]/1.25 [0.12, 12.82]0.03/0.34
Rumbold, 200611118771.15 [0.90, 1.46]NS
Roberts, 201011499691.10 [1.03, 1.17]0.004
aMcCance, 20101167491.02 [0.68, 1.53]NS
Xu, 201023631.04 [0.89, 1.22]NS
Preterm birthaSteyn, 20031212001.43 [1.02, 1.99]d0.031.00 [0.89, 1.12]/1.02 [0.78, 1.34]0.99/0.93
aBeazley, 20051091001.28 [0.75, 2.32]dNS
aCasanueva, 2005731090.55 [0.24, 1.26]dNS
Poston, 200611027481.07 [0.93, 1.22]NS
Rumbold, 200611118771.02 [0.73, 1.43]NS
aVillar, 200911313430.88 [0.74, 1.03]NS
Roberts, 201011499690.97 [0.87, 1.09]NS
aMcCance, 20101167490.83 [0.69, 1.00]0.05
Xu, 201011525361.07 [0.89, 1.29]NS
Low birthweight [<2500 g]aBeazley, 20051091000.46 [0.09, 2.46]dNS0.99 [0.83, 1.19]/1.01 [0.77, 1.33]0.91/0.63
aPoston, 200611027841.16 [0.85, 1.57]NS
aSpinnato, 20081176981.12 [0.60, 2.11]cNS
aVillar, 200911315150.91 [0.80, 1.05]NS
Roberts, 201011497810.93 [0.81, 1.07]NS
Very low birthweight [<1500 g]aPoston, 200611027841.15 [1.02, 1.30]0.021.01 [0.58, 1.73]0.97
aSpinnato, 20081176981.03 [0.71, 1.49]cNS
aVillar, 200911315150.83 [0.59, 1.14]NS
Small-for-gestational age [<10th percentile]aChappell, 19991082830.74 [0.50, 1.09]dNS0.94 [0.82, 1.09]/0.94 [0.76, 1.17]0.35/0.50
aBeazley, 20051091000.99 [0.52, 1.91]dNS
aPoston, 200611027841.10 [0.97, 1.25]NS
Rumbold, 200611118530.87 [0.66, 1.16]NS
aSpinnato, 20081176981.00 [0.72, 1.38]cNS
aVillar, 200911311650.92 [0.75, 1.12]NS
aMcCance, 20101167450.64 [0.39, 1.05]NS
Xu, 201012225360.92 [0.73, 1.15]NS
StillbirthaGulmezoglu, 1997120560.84 [0.36, 1.93]NS1.23 [0.85, 1.94]/1.19 [0.65, 2.15]0.18/0.60
aSteyn, 2003121200N/AeNS
aPoston, 200611027841.65 [0.99, 2.75]d0.06d
Rumbold, 200611118771.34 [0.47, 3.86]NS
aSpinnato, 20081176980.80 [0.30, 2.17]cNS
aMcCance, 20101167611.13 [0.44, 2.91]NS
Xu, 201011525361.73 [0.63, 4.78]NS
Neonatal deathaGulmezoglu, 1997120405.0 [0.64, 39.06]NS0.84 [0.63, 1.13]/0.86 [0.59, 1.25]0.21/0.75
aSteyn, 20031211990.67 [0.11, 3.99]dNS
aPoston, 200611027230.82 [0.40, 1.70]dNS
Rumbold, 200611118630.25 [0.03, 2.24]NS
aSpinnato, 20081176821.16 [0.37, 3.60]cNS
aVillar, 200911315150.83 [0.59, 1.17]NS
Roberts, 201011497810.92 [0.72, 1.19]NS
aMcCance, 20101167300.67 [0.11, 3.99]NS
Xu, 201011525131.73 [0.41, 7.25]NS
OutcomeReference (first author and year)nMean (SD) in gP-valueOverall difference [95% CI]Overall P-value
Intervention/control
  • a

    High pregnancy risk study population.

  • b

    Adjusted odds ratio.

  • c

    Adjusted for study site and pre-enrolment risk group.

  • d

    Values calculated; not reported in text.

  • e

    No stillbirths in the control group.

  • f

    Median.

  • N/A, applicable; NS, not significant; PROM, premature rupture of membranes; RR, relative risk.

BirthweightaChappell, 19991082833100 (N/A)/3160 (N/A)NS−26 g [−82, 29]/−0.07 g [−0.22, 0.09]0.30/0.37
aSteyn, 20031212002491/2664fNS
Casanueva, 2005731093015(513)/3015(629)NS
aPoston, 200611027712901(891)/2967(873)NS
Rumbold, 200611118773392(599)/3386(584)NS
aSpinnato, 20081176983019(780)/3047(756)NS
Roberts, 201011497813247(575)/3244(581)NS
aMcCance, 20101167453435(802)/3355(800)NS

Comments

Public health implications and opportunities for intervention with vitamins B6, B12 and C during pregnancy

Whether decreasing tissue biomarkers of the water-soluble vitamins B6, B12 and C during gestation indicate normal physiological change or depletion and deficiency remains a critical question. The high prevalence of subclinical vitamin B12 deficiency in the general population provides a stronger argument that status of this vitamin may be compromised by increased maternal and fetal demands of pregnancy and that bolstering of reserves may prevent some cases of NTDs.

Supplementation

While there is limited evidence of adverse maternal or fetal outcomes from supplementation with physiological or pharmacological doses of vitamins B6 and C during pregnancy, there is also not a consistent demonstration of benefits and clinical deficiencies are rare. Adequate vitamin B6 status during pregnancy may improve dental health, increase conception rate, and decrease spontaneous abortion rates; however, optimal status and prevalence of inadequacy must be more precisely defined based on such outcomes. Vitamin C deficiency has been implicated in pregnancy complications related to oxidative stress. Results of the current meta-analysis of supplementation with vitamin C during pregnancy (500–1000 mg/day, approximately 7–14 times the RDA) differ slightly from those of similar analyses due to varying inclusion criteria;122,124,125 however, a common conclusion is that supplementation does not reduce the incidence of pre-eclampsia, even in women who are at high risk, and may actually increase the risk of gestational hypertension. Evidence to date fails to provide a strong argument for supplementation with vitamins B6 or C during pregnancy beyond the amount found in routine prenatal vitamin–mineral supplements.

Because there have been no pregnancy supplementation trials with vitamin B12 exclusively, direct risks or benefits cannot be evaluated. However, deficiency of vitamin B12 during pregnancy has been associated with increased NTD risk, and no adverse or teratogenic effects have been demonstrated in women who received therapeutic doses of vitamin B12 during pregnancy. Vitamin B12 deficiency is highly prevalent amongst populations in developing countries where intake of animal source foods is limited. Supplementation of women of childbearing age may reduce the incidence of NTDs, which arise during the first 28 days of gestation, and of neurological degeneration in infants born to mothers with poor status. RCTs designed to evaluate outcomes of vitamin B12 supplementation during the periconceptual period and pregnancy are needed to provide support for this theory.

Fortification

Food fortification, which refers to the addition of micronutrients to processed foods, is a powerful strategy for improving the micronutrient status of a population at reasonable cost if the food vehicle is widely consumed.126 Because vitamins B6 and C are readily available in a varied diet and functional consequences of deficiency in pregnancy are poorly substantiated, fortification has not been considered as a means of reducing pregnancy complications or improving infant and child outcomes. For women of childbearing years, reduction of NTDs and other neurological deficits in the offspring provides an impetus for vitamin B12 fortification.

Research gaps

In assessing the need for interventions with vitamins B6, B12 and C during pregnancy, it is important to be aware of the gaps in existing research and knowledge regarding the risks and benefits of supplementation. Supplementation with pharmacological doses of vitamin B6 during pregnancy has been studied only in conjunction with other antiemetic therapies, most notably doxylamine. Adverse effects of combination treatment have been minimal; however, benefits such as protection from congenital defects cannot be attributed directly to vitamin B6. There is a paucity of follow-up studies of children whose mothers were exposed to physiological or pharmacological doses of vitamin B6 during pregnancy.

No trials published to date have supplemented pregnant women with vitamin B12 exclusively, possibly because severe vitamin B12 deficiency is rare amongst women in wealthier countries and ethical constraints preclude omission of folate supplementation for preventing NTDs. It would be prudent for researchers to investigate maternal, neonatal and child outcomes of supplementation with physiological and pharmacological doses of vitamin B12 in populations at risk for deficiency.

Nearly all trials of vitamin C supplementation during pregnancy have used similar doses of vitamin C (1000 mg/day) and have included supplementation with vitamin E. Despite pharmacological antioxidant doses, these protocols have proven largely ineffective at preventing adverse pregnancy and birth events in a variety of populations studied. Supplementation with yet higher doses or with vitamin C exclusively is not warranted based on current knowledge. However, investigation of the effect of increased gestational antioxidant exposure on childhood outcomes may prove informative.

Conclusion

A critical review of evidence from intervention trials with vitamins B6, B12 and C during pregnancy suggests that while risks of supplementation are minimal, benefits are not clearly delineated. Supplementation with vitamin B6 during pregnancy may reduce symptoms of nausea and vomiting, improve dental health, treat some cases of maternal anaemia, and reduce the incidence of some congenital abnormalities, though additional research is needed to confirm these results. In meta-analysis vitamin B6 supplementation had a significant positive effect on birthweight, though total sample size was small and one trial included supplementation with doxylamine. Vitamin C deficiency has been implicated in pre-eclampsia and PROM; however, interventions to improve status during pregnancy have not systematically reduced the incidence of these complications. In meta-analysis, there was a significant treatment effect of vitamin C and E in increasing the risk of pregnancy-related hypertension; other effects of vitamin C or C and E intervention on maternal and neonatal outcomes, including preterm birth, birthweight, and perinatal morbidity and mortality, were not significant. Data on child health outcomes following maternal supplementation with vitamins B6 and C are not available. Overall, existing evidence does not justify vitamin B6 or C supplementation during pregnancy, though well-designed RCTs investigating potential benefits of vitamin B6 supplementation are warranted. Deficiency of vitamin B12 is highly prevalent in women of reproductive age, especially amongst populations with limited intake of animal source foods. To determine whether improvement of maternal status reduces the incidence of NTDs in the offspring, increases breast milk vitamin B12 content during lactation, and improves infant vitamin B12 status, RCTs with vitamin B12 supplementation during the periconceptual period and pregnancy are necessary.

Acknowledgements

We would like to thank Jan Peerson for assisting with the meta-analyses.

Conflicts of interest

The authors declare no conflicts of interest.

Ancillary