Vitamin B₁₂ status in pregnant women and their infants in South India

Abstract

Background/Objectives:

Vitamin B₁₂ deficiency during pregnancy has been associated with increased risk of adverse perinatal outcomes. However, few studies have investigated the burden and determinants of vitamin B₁₂ status in young infants. This study was conducted to determine the associations between maternal and infant vitamin B₁₂ status.

Subjects/Methods:

Pregnant women participating in a vitamin B₁₂ supplementation trial in Bangalore, India, were randomized to receive vitamin B₁₂ (50 μg) or placebo supplementation daily during pregnancy through 6 weeks postpartum. All women received 60 mg of iron and 500 μg of folic acid daily during pregnancy, as per standard of care. This prospective analysis was conducted to determine the associations between maternal vitamin B₁₂ biomarkers (that is, plasma vitamin B₁₂, methylmalonic acid (MMA) and tHcy) during each trimester with infant vitamin B₁₂ status (n=77) at 6 weeks of age.

Results:

At baseline (⩽14 weeks of gestation), 51% of mothers were vitamin B₁₂ deficient (vitamin B₁₂<150 pmol/l) and 43% had impaired vitamin B₁₂ status (vitamin B₁₂<150 pmol/l and MMA>0.26 μmol/l); 44% of infants were vitamin B₁₂ deficient at 6 weeks of age. After adjusting for vitamin B₁₂ supplementation, higher vitamin B₁₂ concentrations in each trimester were associated with increased infant vitamin B₁₂ concentrations and lower risk of vitamin B₁₂ deficiency in infants (P<0.05). After adjusting for vitamin B₁₂ supplementation, infants born to women with vitamin B₁₂ deficiency had a twofold greater risk of vitamin B₁₂ deficiency (P<0.01). Higher maternal folate concentrations also predicted lower risk of vitamin B₁₂ deficiency in infants (P<0.05). Impaired maternal vitamin B₁₂ status, which combined both circulating and functional biomarkers, was the single best predictor of infant vitamin B₁₂ status.

Conclusions:

Impaired maternal vitamin B₁₂ status throughout pregnancy predicted higher risk of vitamin B₁₂ deficiency in infants, after adjusting for vitamin B₁₂ supplementation. Future interventions are needed to improve vitamin B₁₂ status periconceptionally, and to ensure optimal vitamin B₁₂ status and health outcomes in pregnant women and their children.

Introduction

Vitamin B₁₂ deficiency is a major threat to public health globally.^{1, 2} The prevalence of vitamin B₁₂ deficiency is highest in resource-limited settings, including South Asia.^{3, 4, 5, 6, 7, 8} Vitamin B₁₂ is obtained in the diet through consumption of animal products, including meat, poultry, fish, eggs and dairy. Several studies have reported low vitamin B₁₂ status in vegan or vegetarian individuals and in low- and middle-income settings, particularly in populations with low intake of animal source foods.^{9, 10} In particular, the burden of vitamin B₁₂ deficiency in India is thought to be among the highest in the world.¹

Maternal vitamin B₁₂ deficiency has been associated with greater risk of pregnancy complications, such as spontaneous abortion, low birth weight, intrauterine growth restriction and neural tube defects.¹¹ Children born to women with vitamin B₁₂ deficiency have an increased risk of adverse health outcomes, including deficits in growth and development and anemia.^{12, 13, 14} In the parent-randomized trial in Bangalore, India, daily maternal vitamin B₁₂ supplementation (50 μg/day) with iron and folic acid during pregnancy through 6 weeks postpartum significantly improved maternal vitamin B₁₂ status (P<0.01), breast milk (P<0.01) and infant (P<0.01) vitamin B₁₂ concentrations, compared to iron-folic acid alone.¹⁵

Previous studies in Turkey, Germany, Norway and Brazil have reported associations between maternal and infant vitamin B₁₂ status at birth.^{15, 16, 17, 18} However, few prospective studies have been conducted to examine the burden and determinants of vitamin B₁₂ status in young infants, and there is limited data from India.

In the parent-randomized trial in Bangalore, India, pregnant women were randomized to daily maternal vitamin B₁₂ supplementation with iron and folic acid during pregnancy through 6 weeks postpartum, compared to iron-folic acid alone, to determine the effects on maternal, breast milk and infant vitamin B₁₂ concentrations.¹⁵ We conducted this prospective analysis among 77 mother–infant pairs who were participating in this randomized trial to: (1) determine the prevalence and determinants of inadequate vitamin B₁₂ status during pregnancy and early childhood; and (2) examine the associations of maternal vitamin B₁₂ biomarkers at each trimester with infant outcomes at 6 weeks of age, including vitamin B₁₂, vitamin B₁₂ deficiency, methylmalonic acid and homocysteine concentrations.

Materials and methods

Study population

Participants were pregnant women who were enrolled in a randomized, double-blind, placebo-controlled trial of vitamin B₁₂ supplementation in Bangalore, India. This trial was conducted to examine the effects of daily prenatal vitamin B₁₂ supplementation on biomarkers of maternal vitamin B₁₂ status during pregnancy. The detailed design of the study has been previously described.¹⁵ Briefly, pregnant women were recruited from Hosahalli Referral Hospital in Bangalore, India, and randomized to receive vitamin B₁₂ supplementation (50 μg/day) or placebo daily during pregnancy through 6 weeks postpartum. All women received 60 mg of iron and 500 μg of folic acid supplementation daily beginning at their first prenatal visit, as per standard of care.

Pregnant women were eligible for the study if they were at least 18 years of age, ⩽14 weeks of gestation at enrollment, healthy and carrying a single fetus. Women were excluded if they had any known medical complications, including HIV infection, hepatitis B or syphilis. Women with serious pre-existing medical conditions, previous cesarean section, or who were taking daily vitamin supplements in addition to iron-folate were also excluded. A flow chart of participants in this study is presented in Figure 1.

Figure 1

Summary of enrollment and analysis of samples.

Ethics

The research protocols and study procedures were approved by the Institutional Ethical Board of St John’s Medical College and the TH Chan Harvard School of Public Health Human Subjects Committee. Written informed consent was obtained from all participants. A Data Safety and Monitoring Board met twice annually during the course of the trial.

Follow-up procedures

Structured interviews were conducted to collect information on socio-demographic characteristics, including maternal age, educational level, socioeconomic status and obstetric history. A clinical examination was conducted including vital signs and blood pressure, and obstetric, reproductive and neurological examinations were conducted. Detailed clinical, socio-demographic and anthropometric data were collected prospectively. Maternal weight was recorded using a digital balance to the nearest 100 g; height was measured using a stadiometer to the nearest 0.1 cm; and mid-upper arm circumference, and triceps, biceps and sub-scapular skinfold thickness measurements were measured in triplicate by trained research assistants.

Laboratory investigations: blood sample collection

Maternal blood samples were collected at study visits during each of the three trimesters, or early (<14 weeks gestation), mid- (24 weeks), and late- (34 weeks) gestation (that is, median (interquartile range (IQR)); T1: 10.6 (9.1, 12.6); T2: 24.1 (23.7, 25.0); T3: 33.1 (32.7, 33.6) weeks, respectively), and infant blood samples were collected at 6 weeks of age by venipuncture. The laboratory procedures and biochemical analyses in this trial have previously been described.¹⁵ Briefly, approximately 10 ml of blood was collected from mothers during pregnancy and their infants at 6 weeks of age by venipuncture in both EDTA and plain vacutainer tubes (BD Biosciences, Haryana, India) and stored on ice until centrifugation (<4 h). Whole blood samples were analyzed for hemoglobin and complete blood count, using an automated Coulter counter (ABX Pentra C+; Horiba Medical, New Delhi, India). Plasma and red blood cells were separated and stored at or below −80 °C until analysis for plasma vitamin B₁₂, homocysteine, methylmalonic acid and erythrocyte folate concentrations.

Biomarkers of vitamin B₁₂ status

Plasma vitamin B₁₂ was measured via electrochemiluminescence (Elecsys 2010, Roche Diagnostics, Mannheim, Germany). The intraday and interday assay CVs for plasma vitamin B₁₂ were 0.54 and 2.44%, respectively. Plasma methylmalonic acid and tHcy were assessed by gas chromatography-mass spectrometry (Varian 3800, Palo Alto, CA, USA).¹⁶ The intraday assay CVs for plasma methylmalonic acid (MMA) and tHcy were 6.92 and 5.60%, and the interday assay CVs were 5.57% and 5.04%, respectively. Erythrocyte folate concentrations were determined by a competitive immunoassay with direct chemiluminescence detection on an automatized immunoanalyzer (ADVIA Centaurs, Bayer Health Care Diagnostics, Tarrytown, NY, USA),¹⁷ with intra-assay and interassay variabilities of 1.9 and 5.2%, respectively. The folate concentrations in the hemolysate were converted to whole blood values by adjusting for hematocrit. The laboratory procedures and biochemical analyses are described in further detail in the primary randomized trial.¹⁵

Statistical analyses

Vitamin B₁₂ deficiency was defined as plasma vitamin B₁₂ concentrations less than 150 pmol/l.¹⁹ Impaired vitamin B₁₂ status was defined as plasma vitamin B₁₂ <150 pmol/l plus MMA >0.26 μmol/l. cB₁₂, a combined indicator of vitamin B₁₂ status, modified for three biomarkers (that is, vitamin B₁₂, MMA, tHcy or 3cB₁₂), was calculated using the method and classification developed by Fedosov et al.²⁰ (that is, cB₁₂=_log10[(holoTC*B₁₂)/(MMA*Hcy)]—(age factor)). In Fedosov’s method, the following cutoffs are used to categorize five levels of the combined indicator cB₁₂: probable deficiency (cB₁₂ <−2.5), possible deficiency (−2.5 to <−1.5), low vitamin B₁₂ (−1.5 to <−0.5), vitamin B₁₂ adequacy (−0.5 to <1.5) and elevated vitamin B₁₂ (cB₁₂ ⩾1.5).²⁰

Variables were defined using conventional cutoffs, where available; otherwise, medians of variables were defined based on their distributions in this population. Non-normally distributed variables (that is, plasma vitamin B₁₂, MMA, tHcy, folate concentrations), were natural logarithmically transformed to ensure normality before analysis. Non-transformed values are presented in Tables 1 and 2, for interpretation purposes.

Table 1 Characteristics of the study population

Table 2 Maternal and infant vitamin B₁₂ status

Linear and binomial regression models were used to examine the associations of maternal vitamin B₁₂ biomarkers at each trimester with infant outcomes at 6 weeks of age, including vitamin B₁₂ concentrations (continuous), vitamin B₁₂ deficiency (categorical), methylmalonic acid (continuous) and homocysteine (continuous) concentrations.^{21, 22, 23} Associations between maternal biomarkers of vitamin B₁₂ status from each trimester and infant outcomes were examined independently in separate models. Maternal vitamin B₁₂ supplementation significantly increased maternal and infant plasma vitamin B₁₂ concentrations in the aforementioned randomized trial;¹⁵ therefore, vitamin B₁₂ supplementation regimen was included as a covariate in all models. All models also included an adjustment for the gestational age at sample collection to account for variation in timing of samples. We used the Rothman and Greenland approach to evaluate an extensive list of potential confounders and identify covariates for inclusion in multivariate models, in which all known or suspected risk factors which led to >10% change in effect estimates were included in the model.²⁴ Additional baseline maternal risk factors for infant outcomes were included in multivariate models to evaluate the robustness of the observed associations, including the following: maternal education (⩾10th grade vs <10), standard of living index (⩾28 vs <28), total maternal lymphocyte counts, and maternal body mass index at baseline. The missing indicator method was used to retain observations with missing covariate data.²⁵ We also explored the potential interaction between the randomized intervention and biomarker outcomes, and potential effect modification of observed associations by the randomized intervention. Statistical analyses were conducted using SAS version 9.4 (SAS Institute, Inc., Cary, NC, USA).

Results

The characteristics of participants included in this study are presented in Table 1. Baseline characteristics of pregnant women enrolled in the parent trial and current analysis were similar on age, socioeconomic status and nutritional indicators.

Vitamin B₁₂ status during each trimester of pregnancy and in infants at 6 weeks of age (n=77) are presented in Table 2. At their first prenatal visit, ~51% of pregnant women had vitamin B₁₂ deficiency (vitamin B₁₂<150 pmol/l), 43% had impaired vitamin B₁₂ status (vitamin B₁₂<150 pmol/l plus MMA>0.26 μmol/l) and 38% had low folate status (erythrocyte folate<340 nmol/l). A total of 44% of infants were vitamin B₁₂ deficient and 16% had impaired vitamin B₁₂ status at 6 weeks of age.

The associations between maternal vitamin B₁₂ status in each trimester and infant vitamin B₁₂ concentrations at 6 weeks of age are presented in Table 3. Higher maternal plasma vitamin B₁₂ levels in each trimester (T) were associated with higher vitamin B₁₂ concentrations in infants' multivariate analyses, after adjusting for vitamin B₁₂ supplementation status, gestational age of sample collection, maternal education, standard of living index, lymphocytes and body mass index. Similarly, vitamin B₁₂ deficiency in each trimester was associated with lower vitamin B₁₂ levels in infants in multivariate analyses, after adjusting for vitamin B₁₂ supplementation and other socio-demographic factors. In contrast, higher maternal MMA concentrations predicted lower vitamin B₁₂ concentrations in infants in multivariate analyses. After adjusting for vitamin B₁₂ regimen, impaired maternal vitamin B₁₂ status was associated with lower infant vitamin B₁₂ levels. Higher maternal red blood cell folate levels were also associated with greater vitamin B₁₂ concentrations in infants. In analyses that considered maternal vitamin B₁₂ indicators alone or in combination, impaired maternal vitamin B₁₂ status (that is, vitamin B₁₂ deficiency and elevated MMA) was the strongest and most consistent predictor of infant vitamin B₁₂ status. Findings were similar in both univariate and multivariate analyses, after adjusting for other variables.

Table 3 Associations between maternal vitamin B₁₂ status and infant vitamin B₁₂ concentrations

The associations between maternal vitamin B₁₂ status in each trimester of pregnancy and risk of vitamin B₁₂ deficiency in infants at 6 weeks of age are presented in Table 4. Higher maternal vitamin B₁₂ levels predicted lower risk of vitamin B₁₂ deficiency in infants' multivariate analyses. Infants born to mothers who were vitamin B₁₂-deficient or who had impaired vitamin B₁₂ status had a two to three times greater risk of vitamin B₁₂ deficiency, after adjusting for the vitamin B₁₂ regimen. In contrast, higher maternal MMA concentrations were associated with greater risk of infant vitamin B₁₂ deficiency in multivariate analyses. Higher maternal folate levels were associated with lower risk of vitamin B₁₂ deficiency in infants. Findings were similar in both univariate and multivariate analyses, after adjusting for potential confounders.

Table 4 Associations between maternal vitamin B₁₂ status and infant vitamin B₁₂ deficiency

The associations between maternal vitamin B₁₂ status in each trimester and infant MMA concentrations are presented in Table 5. Vitamin B₁₂ deficiency and elevated MMA concentrations during pregnancy were associated with higher infant MMA concentrations. Impaired maternal vitamin B₁₂ status during pregnancy also predicted higher MMA concentrations in infants. Higher folate levels during pregnancy were also associated with lower infant MMA concentrations.

Table 5 Associations between maternal vitamin B₁₂ status and infant methylmalonic acid

The associations between vitamin B₁₂ status in pregnancy and infant homocysteine levels are presented in Table 6. Higher vitamin B₁₂ and folate levels during pregnancy predicted significantly lower infant tHcy concentrations. Maternal vitamin B₁₂ deficiency and MMA levels were associated with significantly higher infant tHcy concentrations. Maternal impaired vitamin B₁₂ status during pregnancy was also associated with higher tHcy concentrations in infants after adjusting for vitamin B₁₂ regimen, although this was statistically significant in the second and third trimesters. There were no significant associations noted for maternal homocysteine and infant tHcy concentrations.

Table 6 Associations between maternal vitamin B₁₂ status and infant homocysteine

Discussion

In this prospective analysis among pregnant women participating in a vitamin B₁₂ supplementation trial, maternal vitamin B₁₂ status during each trimester significantly predicted vitamin B₁₂ status in infants at 6 weeks of age, even after adjusting for vitamin B₁₂ supplementation. Infants born to mothers who were vitamin B₁₂ deficient (<150 pmol/l) or who had impaired vitamin B₁₂ status (vitamin B₁₂ <150 pmol/l plus MMA >0.26 μmol/l) had higher risk of being vitamin B₁₂ deficient by 6 weeks of age, after adjusting for vitamin B₁₂ regimen. Higher maternal vitamin B₁₂ and folate status, but not maternal homocysteine, were associated with significantly lower infant tHcy concentrations. Impaired maternal vitamin B₁₂ status, which combined both circulating and functional biomarkers, was the single best predictor of infant vitamin B₁₂ status. Higher maternal folate concentrations in pregnancy were also associated with lower risk of vitamin B₁₂ deficiency in infants.

The prevalence of vitamin B₁₂ deficiency was high in this study as follows: 51% of mothers were vitamin B₁₂ deficient and 42% had impaired vitamin B₁₂ status at their first prenatal visit, and 44% of children had vitamin B₁₂ deficiency at 6 weeks of age.¹⁵

Previous studies have noted correlations between maternal and neonatal vitamin B₁₂ status at delivery.^{18, 26, 27, 28} For example, maternal and cord blood holoTC levels were significantly correlated at delivery in a cross-sectional study in Germany (r=0.68, P<0.001).¹⁸ However, most research to date examining the associations between maternal and infant vitamin B₁₂ status have been case–control or cross-sectional in design, and have relied on assessment of a single vitamin B₁₂ biomarker at one time point (for example, maternal vitamin B₁₂ concentrations at delivery), which constrains interpretation of findings.

Maternal vitamin B₁₂ levels during pregnancy are thought to be associated with fetal^{18, 29} and infant³⁰ vitamin B₁₂ concentrations. Some studies have noted significant associations between maternal and neonatal serum vitamin B₁₂ concentrations, whereas prospective cohort studies in India²⁹ and Pakistan³¹ have reported that neonatal vitamin B₁₂ concentrations were 27% to twofold higher than maternal vitamin B₁₂ concentrations. In the current study, infant vitamin B₁₂ concentrations were not significantly different than maternal vitamin B₁₂ concentrations in pregnancy. However, few studies to date have measured maternal vitamin B₁₂ status prospectively throughout the course of pregnancy and examined its association with vitamin B₁₂ status in their infants early in life.

This analysis included a comprehensive assessment of maternal vitamin B₁₂ status prospectively throughout pregnancy, including both circulating (vitamin B₁₂) and functional (MMA, tHcy) vitamin B₁₂ biomarkers. We also included impaired vitamin B₁₂ status and calculated cB₁₂ as a combined indicator of three vitamin B₁₂ biomarkers (that is, vitamin B₁₂, MMA and tHcy), using methods developed by Fedosov et al.²⁰ In analyses that considered maternal vitamin B₁₂ indicators alone or in combination impaired maternal vitamin B₁₂ status (B₁₂<150 pmol/l plus MMA>0.26 μmol/l) was the strongest and most consistent predictor of infant vitamin B₁₂ status. Vitamin B₁₂ biomarkers were also assessed beginning early in gestation (⩽14 weeks), and maternal erythrocyte folate concentrations were assessed prospectively during pregnancy. In addition, the measurement of infant venous blood and comprehensive assessment of infant vitamin B₁₂ status (that is, vitamin B₁₂, MMA and tHcy) were strengths of this analysis.

Our study had several limitations. The assessment of infant vitamin B₁₂ status at a single time point (that is, at 6 weeks of age) and number of infant blood samples available for laboratory analyses (n=77) limit interpretations of the associations between maternal vitamin B₁₂ and infant status early in life. Our findings suggest that participants in the current study were similar to the parent-randomized trial on socio-demographic and nutritional variables; however, they may differ on other unmeasured covariates. Assessment of maternal vitamin B₁₂ status beginning ⩽14 weeks gestation may not reflect periconceptional vitamin B₁₂ status or the relevant etiologic period(s) for vitamin B₁₂ status and perinatal outcomes. The cB₁₂ measure developed by Fedosov et al.²⁰ and modifications for two, three or four biomarkers were constructed based on statistical models in non-pregnant (and primarily elderly) men and women in Chile, Denmark, United Kingdom, Ireland, and the United States. However, cB₁₂ has not been investigated or validated in pregnant women or young infants, which represents a limitation and constrains the interpretation and generalizability of findings. In addition to total vitamin B₁₂, MMA and tHcy, assessment of maternal and infant holotranscobalamin may also represent a better circulating biomarker of vitamin B₁₂ status and transportation from the maternal to the fetal circuit, as vitamin B₁₂ enters and exits the placental villous tissue bound to transcobalamin.³² Vitamin B₁₂ metabolism is also influenced by other nutrients; assessment of infant folate status would further strengthen this analysis. Although findings in this study provide evidence of associations of maternal and infant vitamin B₁₂ status within a randomized trial, the interpretation of these associations is not causal. Future research is needed to elucidate mechanisms of maternal–infant vitamin B₁₂ transport, and the potential role of vitamin B₁₂ in functional outcomes and child health.

In summary, in a large cohort of pregnant women participating in a randomized vitamin B₁₂ supplementation trial in South India, vitamin B₁₂ status throughout pregnancy significantly predicted vitamin B₁₂ status in infants at 6 weeks of age, even after adjusting for vitamin B₁₂ supplementation and several socio-demographic characteristics. Overall, impaired maternal vitamin B₁₂ status, which combined both circulating and functional biomarkers, was the best predictor of infant vitamin B₁₂ status. Infants who were born to women with vitamin B₁₂ deficiency or those with impaired vitamin B₁₂ status had two to four times greater risk of being vitamin B₁₂ deficient, after adjusting for vitamin B₁₂ supplementation status. Findings suggest that although prenatal vitamin B₁₂ supplementation significantly improves vitamin B₁₂ status, maternal vitamin B₁₂ status early in pregnancy has an important role in determining vitamin B₁₂ status early in life. Future research is needed to improve vitamin B₁₂ status in women of reproductive age, and ensure optimal vitamin B₁₂ status and health outcomes in pregnant women and their children.