Introduction
The structure and function of the nervous system of the developing child may be altered permanently by suboptimal maternal nutrition during pregnancy and lactation.Reference Fernstrom 1 , Reference Georgieff 2 Vitamin B12 (B12) and folate deficiencies can interfere in early brain development and function, often by restricting the myelination and synaptic connectivity that occur early in life.Reference Lovblad, Ramelli and Remonda 3 Myelin is the supportive sheath that protects the nerve cells and facilitates transmission of neuronal signals. Synaptic connectivity depends on the quantities of neurotransmitters available and the density of their receptors. If this is altered, neuroanatomical, chemical or metabolic changes may be induced. Functional consequences of these alterations vary depending on the availability of B12 and folate at appropriate times during neurodevelopment.
There are not many reports of associations of maternal B12 and folate status during pregnancy with offspring psychomotor development. Case studies have shown delayed neurodevelopment in infants of vegan mothers and in mothers with pernicious anemia.Reference Dror and Allen 4 In the Mysore Parthenon study, higher maternal folate during pregnancy was associated with better cognitive function in the offspring in childhood.Reference Veena, Krishnaveni and Srinivasan 5 Our earlier Pune Maternal Nutrition Study (PMNS) showed that maternal B12 concentrations during pregnancy were positively related to cognitive performance in the offspring at 9 years (9 y) of age.Reference Bhate, Deshpande and Bhat 6 A Mexican studyReference Del Rio Garcia, Torres-Sanchez and Chen 7 reported that maternal dietary B12 deficiency in the first trimester was associated with poor mental development of children in their first year of life. Folate deficiency was associated with poor cognition in children of mothers who had methyl tetra folate reductase-MTHFR677TT genotype. Studies in rodents have shown that maternal folate deficiency is associated with structural brain abnormalitiesReference Craciunescu, Brown and Mar 8 and poor postnatal learning ability in offspring.Reference Whitley, O'Dell and Hogan 9
In India, where the majority of the population is vegetarian, 10 there is a high prevalence of B12 deficiency and folate status is adequate.Reference Refsum, Yajnik and Gadkari 11 We now report the effect of B12 and folate nutrition during pregnancy on child B12 and folate status and motor, mental and social development at 2 y.
Design and methods
In this study, urban and rural women and their 2-year-old offspring were investigated. Details of the study design were published earlier.Reference Katre, Bhat and Lubree 12
In short, women attending the King Edward Memorial (KEM) Hospital, Pune and the Vadu Rural Health Centre antenatal clinics were enrolled between May 2004 and February 2006 in the study. We screened records of over 995 pregnant women who attended the antenatal clinic before 28 weeks of gestation. Women with a gestation period of longer than 28 weeks, a multiple or an abnormal pregnancy, adverse obstetric history (previous cesarean section, intrauterine growth retardation, fetal death, neonatal death, preeclampsia, twin pregnancy) or a chronic medical condition (diabetes, hypertension, chronic infectious illness) were excluded, and of those remaining (n = 234) agreed to participate. Two hundred and nine of these women delivered at the study centers. Of these, 123 singleton pregnancies with serial measurements at 28 (range 24–30) and 34 (32–36) weeks of gestation, at birth and 2 y follow-up were analyzed (Fig. 1). In addition to demographic and clinical details, blood was sampled at 28 and 34 weeks of gestation for plasma B12 and folate measurements. Socioeconomic status (SES) was measured using the Indian standard of living index (SLI) scale. 13 Nutritional assessment was done by a trained nutritionist. Maternal diet during pregnancy (at 28 and 34 weeks of gestation) and the child's diet at 2 y were assessed using a semi-quantitative food frequency questionnaire based on local practices (investigated through focus group discussion) to obtain the frequency of consumption of foods that contain B12 and folate.
Fig. 1 Flow chart of study participants.
At delivery, the child's weight (to the nearest 0.01 kg) and head circumference (to the nearest 0.1 cm; Pedobaby, ETS JMB, Brussels, Belgium) were measured. At 2 y, offspring were followed up for measurements of hemoglobin, B12 and folate, and motor, mental and social development tests were conducted.
The study was approved by the KEM Hospital Research Centre's ethics committee and informed consent was signed by all participants.
Biochemical measurements
Hemoglobin was measured on a Beckman Coulter Analyzer (AC.T diff ™Analyzer, Miami, Florida, USA). The remaining blood was centrifuged (4°C, 2500 g × 15 min) within an hour of collection, and plasma was stored at −70°C until further analysis. Plasma B12 was measured by a microbiological assay using a colistin sulfate-resistant strain of Lactobacillus Leichmanii.Reference Kelleher and Broin 14 , Reference Kelleher, Walshe, Scott and O'Broin 15 Plasma folate was measured by a microbiological assay using a chloramphenicol-resistant strain of Lactobacillus Casei.Reference Horne and Patterson 16 , Reference Tamura, Freeberg and Cornwell 17
Psychomotor development
For psychomotor assessment, the developmental assessment scale for Indian infants (DASII) was used. DASIIReference Phatak 18 is an Indian adaptation of Bayley's scaleReference Bayley 19 for infants and was administered by a trained psychologist. The motor scale includes 67 items and the mental scale 163 items, which progressively test developmental milestones from birth to 36 months of age. The motor development assessment includes the child's movement from supine to erect posture and locomotion and basic locomotive skills, such as climbing, jumping and skipping. It also includes the record of manipulatory behavior such as reaching, handling and manipulating objects. Mental development items record the child's cognizance of objects in the surrounding environment, perceptual pursuit and exploration to meaningful manipulation. It also covers communication, language comprehension, spatial skills, manual dexterity and social interaction. The number of items achieved is matched to the median motor or mental age for this level for a reference child.Reference Phatak 18 Motor and mental age are converted into motor and mental developmental quotients by the formula: (motor or mental age/chronological age) × 100.
Social development
The revised Vineland Social Maturity ScaleReference Malin and Bharath 20 was used to assess social development. Social maturity is an age-dependent measure of the ability to communicate and interact, and milestones include day-to-day activities such as dressing, drinking, eating with hands and talking in short sentences. The Vineland Social Maturity Scale questionnaire was completed by the researcher during a face-to-face interview of the mother with clarification and further explanation of the questions and answers as necessary. Social age and quotient are derived from the raw scores as for the motor and mental scores.
Statistical methods
Data are presented as arithmetic mean (s.d.) for normally distributed variables, as median (25th, 75th centiles) for skewed variables and as number and percentage (%) for categorical variables. Skewed variables, maternal B12 at 34 weeks, child's B12 and folate at 2 y, were normalized using natural logarithmic transformations. Gender differences were analyzed using Student's unpaired t-test. Correlations between maternal B12, folate and offspring psychomotor and social development are reported as Pearson's or Spearman's correlation coefficients. The prediction of the child's mental and social developmental quotient was by backward regression using the independent variables: birth weight and neonatal head circumference, child's current B12 and folate, separately, maternal B12 and folate status during pregnancy (at 28 and 34 weeks of gestation) adjusted for gender. Maternal age, education, SLI and parity did not have any association with outcomes, and hence were not included in the models reported. This approach was not possible for the motor developmental quotient because of the skewed distribution. A binary group variable for the motor developmental quotient was defined as <115 and ⩾115 was created. Differences between groups were analyzed by Student's unpaired t-test. Statistical analysis was carried out using SPSS 16 (SPSS Inc., Chicago, IL, USA).
Results
Maternal characteristics
Briefly, for the purpose of this report, 123 pregnancies were included in the analysis (70 rural, 53 urban) who had all measurements available at 28 (range 24–30) and 34 (32–36) weeks of pregnancy and for the child at 2 y follow-up. Women included in the analysis (n = 123) were not different (P > 0.05) in weight, height, hemoglobin at 34 weeks of gestation and SES compared with those not included (n = 86; data not shown).
As a group, the mothers were young, short and light (Table 1). At 28 weeks of gestation, 78% women had low hemoglobin (<110 g/l), 62% had low B12 concentrations (<150 pmol/l) and only one woman had a low folate concentration (Table 1).
Table 1 Maternal characteristics at 28 and 34 weeks and 2 y after delivery (n = 123)
SES, socioeconomic status; GLV, green leafy vegetables; Hb, hemoglobin.
aMeasured as the Indian standard of living index, 13 values are median (25th–75th centiles), mean ± s.d., n (%).
bMore than twice a week.
Offspring characteristics
There were 61 boys and 62 girls: 78 were first born (63%), 7 (6%) were delivered preterm and 18 (15%) by the cesarean section. Twenty five percent (31) had low birth weight (<2.5 kg).
New born were small and light (Table 2). At 2 y (2.00, 2.04), the children were 1.2 standard deviation scores (SDS) lighter and 0.8 SDS shorter compared with the World Health Organization reference child. 21 Seventy eight (63.4%) children had low hemoglobin (<110 g/l).
Table 2 Characteristics of children at birth and 2 y (n = 123)
Hb, hemoglobin.
Values are median (25th, 75th centiles) or mean ± s.d.
aDifferent by gender, unpaired t-test (P < 0.05).
bDifferent by gender, unpaired t-test (P < 0.01).
Maternal B12 at 28 and 34 weeks of gestation was associated with offspring B12 at 2 y (r = 0.29; r = 0.32; P < 0.001), but maternal folate was not associated with offspring folate. At 2 y, median motor and mental development quotients were 115.4 (95% CI 92.4, 115.4) and 100.0 (95% CI 81.3, 117.9), respectively, and social quotient of the offspring was 94.5 (95% CI 77.1, 100.0). Compared with Indian norms,Reference Phatak 18 the average motor development quotient was on the 84th percentile and mental development quotient on the 50th percentile. Social quotient reference values are not available; a value of more than 100 would be considered ‘advanced’. There was no gender difference for any of the developmental quotients. The only effect of maternal age, education or SES on developmental quotients was that children with a lower SES had a lower motor quotient than children with a higher SES (114.1 v. 115.4, P = 0.006, Mann–Whitney U-test). The urban children performed better than the rural children on motor development quotient (113.7 v. 110.8, P < 0.01, Mann–Whitney U-test).
Motor development associations
We found that at 2 y, motor development quotient was not associated with any of the child's current or birth measurements. It was associated with maternal folate at 28 and 34 weeks of gestation (Spearman's r = 0.29; P = 0.001 and r = 0.22; P = 0.016) but not with B12. Motor development quotient was reclassified into a binary variable (motor development quotient <115 and ⩾115). The only difference found between these groups was that pregnancy folate at 28 and 34 weeks was significantly high in the higher scoring children (Table 3).
Table 3 Child motor development quotient group (low and high) at 2 y and differences in maternal vitamin B12 and folate concentrations at 28 and 34 weeks of pregnancy
aUnpaired Student's t-test, mean ± s.d. and median (95% confidence interval).
Mental development associations
Two-year mental development quotient was not associated with any of the measurements at 2 y or at birth, but was positively associated with maternal B12 at 28 weeks (r = 0.19, P = 0.041). Mental development quotient in children of B12 sufficient mothers (>150 pmol) was higher than that in the children of B12 deficient mothers (101 v. 98, P = 0.035). The child's mental development quotient was also associated positively with head circumference at birth but negatively with birth weight (Table 4). In backward regression models, the mental development quotient was positively and separately associated with maternal B12 at 28 weeks of gestation and folate at 28 and 34 weeks of gestation. A maximum of 12% of the variation was explained by B12 at 28 weeks of gestation (Table 4, model 1).
Table 4 Determinants of child's mental development quotient at 2 y (n = 123)
stdβ, standardized beta.
Models determined by backward regression. The effect of adding maternal vitamin B12 and folate at 28 and 34 weeks pregnancy is separately explored in models 1–4.
Model 1: gender, child B12 at 2 y, birth weight, neonatal head circumference, maternal B12 at 28 weeks.
Model 2: gender, child B12 at 2 y, birth weight, neonatal head circumference, maternal B12 at 34 weeks.
Model 3: gender, child folate at 2 y, birth weight, neonatal head circumference, maternal folate at 28 weeks.
Model 4: gender, child folate at 2 y, birth weight, neonatal head circumference, maternal folate at 34 weeks.
Social development
Two-year social development quotient was not associated with the child's current or birth measurements, but it was positively associated with maternal B12 at 28 weeks of gestation, folate at both 28 (r = 0.18, P = 0.05) and 34 weeks of gestation (r = 0.21, P = 0.02) and the child's hemoglobin at 2 y (r = 0.34; P < 0.001). Social development quotient in children of B12 sufficient mothers (>150 pmol/l) was higher than that in the children of B12 deficient mothers (93 v. 91, P = 0.029). Girls performed better than boys on social development. In backward regression (Table 5), social development was predicted positively by hemoglobin and negatively by child B12 at 2 y. Social development was also associated with birth measurements. Positive association was found with head circumference and negative association with birth weight. On maternal parameters, there was a positive association with maternal B12 at 28 (Table 5, model 1) and 34 (Table 5, model 2) weeks of gestation. Girls performed better than boys in folate models as well, and social developmental quotient was associated negatively with birth weight, positively with head circumference at birth and maternal folate at 28 (Table 5, model 3) and 34 weeks of gestation (Table 5, model 4). If child hemoglobin at 2 y was entered into the folate models, folate was no longer a predictor.
Table 5 Determinants of child's social quotient at 2 y (n = 123)
Hb, hemoglobin; stdβ, standardized beta.
Backward regression, stdβ. The effect of adding maternal vitamin B12 and folate at 28 and 34 weeks pregnancy is separately explored in models 1–4.
Model 1: gender, child B12 at 2 y, birth weight, neonatal head circumference, maternal B12 at 28 weeks.
Model 2: gender, child B12 at 2 y, birth weight, neonatal head circumference, maternal B12 at 34 weeks.
Model 3: gender, child folate at 2 y, birth weight, neonatal head circumference, maternal folate at 28 weeks.
Model 4: gender, child folate at 2 y, birth weight, neonatal head circumference, maternal folate at 34 weeks.
Discussion
Our study demonstrates, in women with low B12 status and adequate folate, a positive association between maternal B12 and folate status during pregnancy with offspring motor, mental and social development at 2 years of age. Our previous study (PMNS) found that lower maternal B12 during pregnancy was associated with lower scores for sustained attention and short-term memory in offspring at 9 years of age.Reference Bhate, Deshpande and Bhat 6 The present study adds to the evidence that maternal B12 insufficiency is associated with lower scores in mental development on the DASII scale. In addition, it demonstrates that motor development at 2 y was predicted by maternal folate during pregnancy. These findings are independent of child's B12 and folate status at 2 y.
Social development quotient was positively associated with hemoglobin at 2 y and maternal B12 at 34 weeks, but negatively with child B12 at 2 y. This is in agreement with a recent reportReference Eilander, Muthayya and van der Knaap 22 of cognitive measures in 6- to 10-year-old children, where a positive association with hemoglobin and a negative association with current B12 was shown. We extend this finding to demonstrate the favorable influence of pregnancy B12 on child development measured at 2 y. We also showed that at 2 y head circumference at birth was positively associated with mental and social development, which may reflect the effect of ‘head-sparing’ in these relatively small offspring. In other words, as the degree of growth retardation increases, head circumference proportionally increases and this increase in size may be associated with higher levels of mental and social development.Reference Kramer, McLean, Olivier, Willis and Usher 23 Our finding that maternal B12 status predicts offspring B12 status at 2 y is confirmed by other studies of B12 deficiency in infants.Reference Jones, Ramirez-Zea, Zuleta and Allen 24 , Reference Schulpis, Spiropoulos and Gavrili 25 Maternal B12 concentrations in our study were comparable to those in other studies in India,Reference Muthayya, Dwarkanath and Mhaskar 26 although lower than those in other populations that are predominantly nonvegetarian.Reference Chambers, Obeid and Refsum 27 Low maternal B12 levels can result in low levels in infantsReference Muthayya, Dwarkanath and Mhaskar 26 with resultant poor brain developmentReference Molloy, Kirke, Brody, Scott and Mills 28 and cognitive impairment. Case studies have shown that maternal B12 deficiency is associated with offspring neurological damage.Reference Kuhne, Bubl and Baumgartner 29 , Reference Stollhoff and Schulte 30 A recent study found that lower folate status in early pregnancy might impair fetal brain development and can be associated with hyperactivity/inattention and peer problems in childhood.Reference Schlotz, Jones and Phillips 31
To our knowledge, ours is one of very few studies that have examined both maternal B12 and folate status and their association with early development of the offspring. Adolescents from macrobiotic families in the Netherlands were reported to have impaired cognitive performance.Reference Louwman, van Dusseldorp and van de Vijver 32 In these children, B12 deficiency was associated with a suboptimal performance on intelligence tests. B12 deficient schoolchildren in Guatemala had slower reaction times on psychological tests, lower academic performance, attention problems and delinquent behavior.Reference Allen, Penland, Boy, DeBaessa and Rogers 33 The Mysore Parthenon study found higher maternal folate and not B12 during pregnancy to be associated with better cognitive function in 536 offspring at 10 years of age,Reference Veena, Krishnaveni and Srinivasan 5 whereas our study finds an association with maternal B12 as well as folate. Direct comparison of our study with other studies is difficult due to different age groups, sample size and measures used.
The possible mechanism linking B12 and folate with psychomotor development is the critical role that both B12 and folate have in one carbon metabolism: methylation, nucleotide synthesis and cell division and differentiation.Reference Mason 34 The shared metabolism means that deficiencies in one vitamin will alter the metabolism of the other.Reference Black 35 In nervous tissue, under conditions of low B12 status, this results in the accumulation of methylmalonyl CoA and the formation and incorporation of nonphysiologic fatty acids into neuronal lipids including those in myelin sheaths.Reference Dror and Allen 4 , Reference Black 35 Neurotransmission may also be affected because methylcobalamin is the obligate carrier of methyl groups for the synthesis of choline, the precursor of the neurotransmitter acetylcholine.Reference Combs 36
The strength of our study is its prospective design and long-term follow-up. Other factors known to affect infant motor and mental development are maternal education, SES, length of gestation, birth weight and head circumference, and these were included in the multivariate analysis. The limitations of our study include lack of data on maternal cognitive measures and availability of the relevant data only on a part of the population.
Early development of competent motor skills, perception and cognition are vital for the future success of the child. Our study provides evidence that maternal B12 and folate concentrations influence offspring brain development and function in addition to increased risk for cardiovascular disease and type 2 diabetes that has been previously reported by our group.Reference Yajnik, Deshpande and Jackson 37 Ideally, women should conceive with adequate B12 and folate status and this should be maintained throughout pregnancy and breast feeding. In India, B12 deficiency is common, and given its association with neural tube defectsReference Godbole, Deshmukh and Yajnik 38 several experts call for B12 fortification of the food supply. However, the interdependence of folate and B12 status on the methylation of DNA with potential lifelong epigenetic programming effects means that an intervention study needs to be designed and carefully evaluated before national fortification can be recommended. Before the introduction of food fortification of B12, advice to women of child-bearing age should be to increase their intake of B12 foods such as animal origin foods, including milk, and consider consumption of supplements particularly before and during pregnancy and lactation.
Acknowledgments
The International Atomic Energy Agency, Vienna, Austria, provided financial support. We are grateful to the pregnant women and their children for taking part in this study. We thank Dr K. J. Coyaji, Director of the KEM Hospital, Pune, for providing research facilities. We thank Dr S. S. Naik, Ms Neelam Memane, Ms Pallavi Hardikar, Ms Deepa Raut, Ms Trupti Konde, Dr Niranjan Joshi, Dr Aparna Joshi, Dr Madhumati Otiv, Mrs P. C. Yajnik, Mr Charudatta Joglekar, Ms L. V. Ramdas, Mr T. M. Deokar, Mr S. D. Chougule, Mrs Vaishali Deshpande, Mr A. B. Gaikwad, and the obstetricians and drivers for their invaluable contributions to the study.
None of the authors had any financial or personal conflicts of interest associated with this manuscript.
