European Journal of Clinical Nutrition (2014) 68, 870–875 © 2014 Macmillan Publishers Limited All rights reserved 0954-3007/14 www.nature.com/ejcn

MATERNAL AND PEDIATRIC NUTRITION HIGHLIGHTS ORIGINAL ARTICLE

Preventing vitamin B12 deficiency in South Asian women of childbearing age: a randomised controlled trial comparing an oral vitamin B12 supplement with B12 dietary advice GJ Mearns1, J Koziol-McLain2, V Obolonkin3 and EC Rush4 BACKGROUND/OBJECTIVES: To examine the effectiveness, acceptability and sustainability of interventions to reduce vitamin B12 (B12) deficiency in South Asian women before conception. SUBJECTS/METHODS: A 6-month randomised controlled trial conducted in Auckland, New Zealand. Participants (62 South Asian women, 18–50 years old) were stratified by dietary practices, then randomised to three treatment groups: B12 Supplement (oral cyanocobalamin 6 μg/day) (n = 21), Placebo (n = 21), or B12 Dietary Advice (n = 20). Primary outcome measures were changes in B12 biomarkers (serum B12 and holotranscobalamin (holoTC)) at 6 months. Dietary B12 intake was estimated from a B12 food-specific frequency questionnaire (B12FFQ). Intention-to-treat analysis was applied using ‘last observation carried forward’ method. Changes in B12 biomarkers by treatment were compared using analysis of variance. Pearson’s correlations tested relationships between dietary B12 intake and B12 biomarkers. RESULTS: At baseline, 48% of women tested as insufficient or deficient in serum B12 ( o222 pmol/l) and 51% as insufficient or deficient in holoTC ( o45 pmol/l). B12 status was moderately correlated with dietary B12 intake (r = 0.5, 95% confidence interval (CI) (0.3–0.7)) and 44% of women reported insufficient dietary intake (o 2.4 μg/day). B12 Supplement was the only treatment group to record a significant increase in B12 biomarkers over 6 months: serum B12 by 30% (95% CI (11–48%)) and holoTC by 42% (12–72%). CONCLUSIONS: The prevalence of B12 insufficiency among Auckland South Asian women is high and moderately correlated with inadequate intake of foods that contain B12. Cyanocobalamin supplementation (6 μg/day) was associated with improved B12 biomarkers, with a potential to improve preconception B12 status in South Asian women. European Journal of Clinical Nutrition (2014) 68, 870–875; doi:10.1038/ejcn.2014.56; published online 16 April 2014

INTRODUCTION Adequate dietary supply and stores of vitamin B12 (B12; a methyl donor in one-carbon metabolic pathways) are essential across the life-course for normal growth, development and function.1–4 B12 deficiency is considered rare in New Zealand (NZ), with public funding for B12 supplements,5 decisions on folic acid supplementation6,7 and management of B12 deficiency8 all orientated towards addressing B12 deficiency in older adults with B12 malabsorption. In India, between 40–75% women of childbearing age have low B12 concentrations,1,3,9 with evidence of functional B12 deficiency (increased concentrations of homocysteine and methylmalonic acid).1,9 The deficiency is linked to intergenerational dietary patterns that are low or absent in animal source foods; patterns historically influenced by culture, religion, geographic origins, family/ community customs and food availability.4,10 Maternal B12 deficiency may contribute to persistent metabolic or functional abnormalities in offsprings across their life-course.11,12 The longitudinal Pune Maternal Nutrition Study in India found associations between maternal B12 deficiency during pregnancy and low muscle mass with increased insulin resistance and relative adiposity in offsprings at 6 years of age.1 These effects were more pronounced when

mothers were low in B12 but high in folate. They add to the already recognised offspring risks associated with B12 deficiency---poorer pregnancy outcomes,13,14 impaired cognition2 and increased incidence of neural tube defects.15,16 There has been little attention to B12 deficiency in NZ women of childbearing age. One NZ study of 124 migrant Indian women (mean age 39.8 ± 9.8 years) found that 59% with vegetarian and 38% with non-vegetarian dietary practices tested as either deficient (o 150 pmol/l) or marginally deficient (150–221 pmol/l) for serum B12 concentrations.17 Strategies to assess and improve B12 status in young South Asian women before conception need to be investigated. The inclusion of a wide variety of nutritious foods in the diet is the preferred public health advice to prevent micronutrient deficiencies, rather than supplementation with vitamins.18 Two focused strategies to improve B12 status are B12 food-specific dietary advice or low-dose supplementation. Better understanding is needed of how tailored dietary advice compares with an equivalent low-dose oral B12 supplement for improving B12 status in women with insufficient dietary B12 intake. The purpose of this study was to identify how B12 status could be improved over 6 months in a group of South Asian women of childbearing age by using either tailored dietary advice or a daily B12

1 School of Health Care Practice, Faculty of Health and Environmental Sciences, AUT University, Auckland, New Zealand; 2Interdisciplinary Trauma Research Centre, Faculty of Health and Environmental Sciences, AUT University, Auckland, New Zealand; 3Centre for Child Health Research, Faculty of Health and Environmental Sciences, AUT University, Auckland, New Zealand and 4Centre for Child Health Research, Faculty of Health and Environmental Sciences, AUT University, Auckland, New Zealand. Correspondence: Dr GJ Mearns, School of Health Care Practice, Faculty of Health and Environmental Sciences, AUT University, Private Bag 92006, Auckland 0630, New Zealand. E-mail: [email protected] Received 25 October 2013; revised 20 February 2014; accepted 28 February 2014; published online 16 April 2014

Vitamin B12 deficiency in childbearing women GJ Mearns et al

supplement capsule (6 μg). The hypothesis was that dietary advice would be more effective than a daily B12 supplement or placebo. The primary outcome was change in B12 biomarker concentrations over 6 months. Other outcomes were sustainability and acceptability of the treatment. SUBJECTS AND METHODS Study design This was a three-arm randomised controlled trial to compare treatments for improving B12 biomakers over 6 months. The three treatment arms were: 6 μg cyanocobalamin supplement daily, placebo daily, or B12 dietary advice. A sample size of 62 particpants was determined to detect a mean change in serum B12 of 66 ± 71 pmol/l over 6 months (attrition rate of 20%, 80% power, α level of 5%), assuming the same effect as that seen for the daily provision of milk (0.96 μg B12 in 200 ml) to 120 children in Kenya.19 The trial was registered with the Australian New Zealand Clinical Trials Registry (ACTRN12610000262000). Ethical approval was obtained from the AUT University Ethics Committee. Six focus group interviews undertaken with South Asian community members and health professionals before commencing the clinical trial provided inputs on community priorities for managing B12 deficiency and guidance for culturally appropriate conduct of the trial.

Participants Sixty-two South Asian women aged 18–50 years living in Auckland, NZ participated in the 6-month trial. Non-random participant recruitment was used; participants self-selected to be a part of the study by responding to recruitment advertisements in local news media or learning of the study through South Asian community contacts. Participants responded to a screening question and were subsequently assigned to one of the two strata based on either meat-eating or non-meat-eating dietary practices. A third party person not directly involved with the randomised controlled trial randomly allocated participants from each strata into a research treatment group using a stratified random assignment process determined by a sequence-matched random number allocation table. The Placebo and B12 Supplement groups were double blinded, but it was not feasible to blind the Dietary Advice group. Exclusion criteria included women with major health conditions, malabsorption disorders, pregnancy or breastfeeding and consumption of medicines, such as metformin, proton pump inhibitors, B12 injections or B12 supplements (Figure 1).

871 Procedure At baseline, medical history and demographic information, age, ethnicity, religion, occupation, number and ages of children and duration of residence in NZ were recorded. Fasting blood samples, anthropometric measurements and estimations of dietary B12 intake (B12 food-specific frequency questionnaire (B12FFQ)) were collected at baseline and at 2 and 6 months. Participants fasted overnight for 12 hours before blood test sampling. Venous blood (15–20 ml) was drawn from the antecubital fossa. The primary outcome measures were serum B12 and holotranscobalamin (holoTC) for B12 status. Full blood count was also measured with ferritin, and iron studies were requested if iron deficency was suspected. Serum B12 and folate were measured using the Modular Analytics E170 immunoassay analyser (ISO accredited laboratory Diagnostic MedLab, Auckland, NZ). The samples for holoTC were collected in a heparin tube (4 ml) and transported on ice before being centrifuged at a university laboratory for 10 min at high speed (5800 r.p.m.) using the Z150A compact lab centrifuge (Labnet International Inc., Edison, NJ, USA). Plasma was aliquoted into storage tubes, placed temporarily in a − 4 °C freezer, before being transferred to a − 80 °C freezer within 4 weeks. Samples were then forwarded in two batches to the local health board laboratory for measurement of holoTC using AxSYM Active-B12 microparticle enzyme immunoassay analyser (Abbott Laboratories, Abbott Park, IL, USA). An internal control sample of plasma was forwarded in both batches to measure inter-batch consistency. A 30-item, researcher-administered food frequency questionnaire, developed for the VitB12 study, was used to estimate dietary B12 intake at baseline and at 2 and 6 months. The B12FFQ grouped foods into dairy, red meat, white meat and seafood, baked goods and fortified food categories. Participants recalled the type, quantity and frequency of these foods consumed over the previous 3 months. The nine different frequencies of consumption of a particular food ranged from ‘never’, ‘one time per month or less’ through to ‘six or more times per day’. These were used to calculate average daily dietary B12 intake. Participants in the Placebo and B12 Supplement groups were advised to swallow either one cyanocobalamin or placebo capsule with a full glass of water each day for 6 months. Repeat bottles of capsules were delivered to participants at 2 and 4 months. Sustainability and adherence to the treatment were assessed from the number of capsules returned over the 6-month period. Dietary Advice group participants received verbal and written guidelines on how to increase or maintain consumption of B12-containing foods in order to meet the 2.4 μg minimum recommended daily intake (RDI) for B12.18 The individualised guidelines were based on food preferences recorded in the participant’s B12FFQ collected at baseline. Sustainability

Enrollment

Assessed for eligibility (n= 75 )

Excluded (n= 13). Reasons: • previous B12 injections (n= 8) • on metformin (n= 3) • thyroid disease (n= 2)

Stratified by current meat-eating (ME) and non meat-eating (NME) dietary practices (n=62)

Analysis

Follow-Up

Allocation

Random allocation ME strata (n= 9)

Figure 1.

Random allocation NME strata (n= 12)

Random allocation ME strata (n= 8)

Random allocation NME strata (n=13)

Random allocation ME strata (n= 10)

Random allocation NME strata (n= 10)

Allocated to Placebo group (n= 21) • Received allocated treatment • (n= 19) • Did not receive allocated treatment: reason- abnormal blood test results (n=2)

Allocated to B12 Supplement group (n=21) • Received allocated treatment • (n= 21) • Did not receive allocated treatment • (n= 0)

Allocated to Dietary Advice group (n= 20) • Received allocated treatment • (n= 20) • Did not receive allocated treatment • (n= 0)

Lost to follow-up:(n=3). Reasons: • Self-withdrawal from study: reasonpregnancy (n=1) • Did not respond to follow-up appointment requests (n=1) • Discontinued intervention - received B12 injection from own doctor (n=1)

Lost to follow-up(n=1). Reasons: • Self-withdrawal from study: reason not given (n=1)

Lost to follow-up (n=5). Reasons: • Self-withdrawal from study: reasontoo busy (n=1) • Did not respond to follow-up appointment requests (n=1) • Discontinued intervention - received B12 injection from own doctor (n=3)

Analysed (n=19). Includes: • Imputed data using last results carried forward (n=3)

Analysed (n=21). Includes: • Imputed data using last results carried forward (n=1)

Analysed (n=20). Includes: • Imputed data using last results carried forward (n=5)

CONSORT50 diagram of participant flow through the VitB12 randomised controlled trial.

© 2014 Macmillan Publishers Limited

European Journal of Clinical Nutrition (2014) 870 – 875

Vitamin B12 deficiency in childbearing women GJ Mearns et al

872 and adherence were assessed through the B12FFQ sampled at the 2- and 6-month data collection visits. Participants were contacted during the week following the implementation of all treatments and again after the 2-month data collection. Where indicated for the Dietary Advice group, the dietary guidelines were modified or reiterated. Acceptability for all treatments was determined from participant reports of experiences and any adverse effects.

Table 1.

Selected baseline characteristics by treatment group Placebo (n = 20)

Dietary B12 advice (n = 19)

P

34 (9.7) (21, 49)

37 (9.7) (18, 50)

0.18

Weight (kg) 66.4 (11.1) (46.3, 87.0)

63.7 (13.5) (46.7, 103.4)

62.9 (10.6) (47.3, 87.2)

0.56

Height (cm) 159 (4.6) (151, 164)

159 (6.1) (144, 171)

160 (5.0) (151, 169)

0.99

29 (8.9) (15, 45)

26 (7.9) (12, 40)

0.22

Hb (g/l) 128 (13.7) (93, 145)

125 (11.4) (99, 144)

132 (7.05) (119, 142)

0.09

RBC ( × 10e 12/l) 4.7 (0.26) (4.3, 5.3)

4.8 (0.43) (4.2, 6.0)

4.6 (0.25) (4.2, 5.0)

0.76

0.39 (0.03) (0.33, 0.44)

0.40 (0.03) (0.32, 0.45)

0.52

MCHC (pg) 27.2 (2.6) (22, 30)

26.3 (3.2) (20, 32)

28.5 (1.6) (25, 31)

0.04a

MCV (fl) 83.8 (5.5) (74, 93)

81.8 (8.1) (64, 93)

86.5 (4.0) (74, 92)

0.07

WBC ( × 10e 9/l) 6.6 (1.7) (3.9, 11.6)

7.0 (1.9) (3.6, 1.6)

6.7 (1.6) (4.2, 9.5)

0.78

B12 supplement (n = 21) Age (years) 39 (7.3) (25, 50)

Statistical analysis Predictive Analytic Software (PASW) Edition 18 (IBM, Chicago, IL, USA) was used to conduct statistical analyses and for graphical representation of data. Missing data for measures of B12 markers and dietary B12 intake were included in primary analyses using intention to treat where the last measurement or test result was carried forward to the next measurement point.20 The 95% confidence intervals (CIs) on Pearson’s correlation coefficients and analysis of variance were calculated using a Microsoft Excel algorithm.21 Literature-derived cutoffs were used to define deficient (serum B12o150 pmol/l, holoTCo35 pmol/l),22–24 insufficient (serum B12 150–221 pmol/l, holoTC 35–44 pmol/l ) or sufficient B12 status (serum B12>221 pmol/l, holoTC>44 pmol/l).15,24,25 Baseline characteristics were compared between treatment groups using analysis of variance for normally distributed data and Kruskal–Wallis for skewed data (Table 1). Skewed data (B12 biomarkers and B12 dietary intake) were log-transformed before further analysis. The effect of group treatment on the change in B12 biomarkers was tested using analysis of covariance with simple planned contrasts to determine the influential group treatment. Pearson’s correlation coefficient tests were used to explore associations between the main outcome measures (change in B12 biomarkers) and other potentially biologically related continuous variables. The three variables with a moderate (r ⩾ 0.5) correlation (age, dietary B12 intake and capsule adherence) to the main outcome measures were included as covariates in the analysis of covariance, and the Bonferroniadjusted significance was determined.26 At baseline, Pearson’s correlations were used to measure associations between estimated dietary intake from the B12FFQ and B12 biomarkers. The adjusted coefficient of determination (R2) was used to quantify the magnitude of association between variables. Changes in B12 biomarkers and dietary B12 intake are reported as a comparison in percentage difference over time using the exponential of the log-transformed mean and 95% CI.27

RESULTS There was a 15% attrition rate, with 51 participants completing 6 months of treatment (Figure 1). The groups were similar for baseline characteristics except for holoTC, which was higher for the Dietary Advice group than for the B12 Supplement and Placebo groups (P = 0.03; Table 1). At baseline, 48% of the participants were insufficient or deficient in serum B12 and 51% were insufficient or deficient in holoTC. All had sufficient folate status (27 ± 8.1). No participant had macrocytic anaemia (haemoglobin (Hb) ⩽ 120 g/l, mean corpuscular volume ⩾ 99f/l). Microcytosis (mean corpuscular volume ⩽ 80 f/l) was present in 10 participants, and follow-up serum ferritin tests revealed iron deficiency in eight of these. The remaining two were unexplained. At baseline B12FFQ and serum B12 measurements, B12 status was moderately correlated with dietary B12 intake (r = 0.5, 95% CI (0.3–0.7)) and 44% women reported insufficient dietary intake (o 2.4 μg/day). The women (n = 22) who reported eating both white and red meat had almost double the serum B12 and holoTC concentrations compared with those who did not (Table 2). These participants were the only group to report a median dietary B12 intake at or above the RDI. Change in B12 biomarkers among treatment groups The B12 supplement treatment was associated with substantial increases in both serum B12 (geometric mean increase of 30%, 95% CI (11–48)) and holoTC (42% (12–72)) over 6 months (Figure 2). The mean change in B12 biomarkers, when compared between treatment groups, was significant for the B12 Supplement group only (serum B12, P = 0.001–0.005; holoTC, European Journal of Clinical Nutrition (2014) 870 – 875

Serum folate (pmol/l) 25 (7.7) (11, 39)

PCV (%) 0.40 (0.03) (0.32, 0.43)

Median (25th, 75th) HoloTC (pmol/l) 30 (21, 58) Serum B12 (pmol/l) 198 (165, 245)

Median (25th, 75th) Median (25th, 75th)

P

47 (31, 85)

58 (36, 76)

0.03b

230 (156, 396)

267 (192, 414)

0.15

Abbreviations: Hb, haemoglobin; HoloTC, holotranscobalamin; MCHC, mean corpuscular haemoglobin concentration; MCV, mean corpuscular volume; PCV, packed cell volume; RBC, red blood cell; WBC, white blood cell. Normally distributed values reported as mean ± s.d. (range). Non-normally distributed values reported as median (25th/75th percentiles). aSignificant difference between Dietary Advice and other groups. bSignificant difference between B12 Supplement and other groups.

P = 0.001–0.018). There was no significant change for the Dietary Advice or Placebo groups (serum B12, P>0.99 and holoTC, P = 0.60). When change in dietary B12 intake was controlled for, group treatment predicted 36% of the change in serum B12 and 22% of the change in holoTC (Table 3). When supplement adherence was taken into account (Placebo and B12 Supplement groups only (n = 39)), group treatment predicted 38% of the change in serum B12 and 22% of the change in holoTC (Table 3). © 2014 Macmillan Publishers Limited

Vitamin B12 deficiency in childbearing women GJ Mearns et al

873 Table 2.

Descriptive baseline B12 biomarkers by dietary practice

category Dietary practice category

n Median

25th percentile

75th percentile

202b 165b 216b 375 229

149c 145c 148c 230 164b

265 276 236 456 365

33d 30d 28d 70 44e

25d 28d 19d 55 29d

a

Serum vitamin B12 Lactovegetarian 26 Lactoovovegetarian 7 White meat eating only 5 White and red meat eating 22 All participants 60

a

HoloTC Lactovegetarian 26 Lactoovovegetarian 7 White meat eating only 5 White and red meat eating 22 All participants 60

60 49 44e 94 70

a

Abbreviation: HoloTC, holotranscobalamin. Serum B12 and holoTC measured in pmol/l, Tukey Hinges interquartile ranges. bSerum B12 insufficient (150–221 pmol/l). cSerum B12 deficient (o150 pmol/l). d HoloTC deficient (o 35 pmol/l). eHoloTC insufficient (35–44 pmol/l).

Table 3. ANOVA and ANCOVA of change in serum B12 and holoTC over 6 months F Serum B12 Serum B12 Serum B12 with CV age Serum B12 with CV dietary B12 Serum B12 with CV capsule adherence HoloTC HoloTC HoloTC with CV age HoloTC with CV dietary B12 intake HoloTC with CV capsule adherence

95% CIa (LL, UL)

P

R2

8.9 7.2 10.2 19.3

(3.8, (2.8, (4.3, (8.2,

14.0) o0.001b 11.5) 0.002c 16.0) o0.001b 30.4) o0.001b

0.21 0.25 0.36 0.38

7.4 6.3 6.9 11.5

(3.1, (2.3, (2.6, (4.5,

12.0) o0.001b 11.0) 0.003d 11.2) 0.002c 18.5) 0.002c

0.18 0.17 0.22 0.22

Abbreviations: ANCOVA, analysis of covariance; ANOVA, analysis of variance; CI, confidence interval; CV, covariate; holoTC, holotranscobalamin; LL, lower limit; R2, adjusted coefficient of determination; UP, upper limit. ANOVA/ANCOVA on log-transformed change in serum B12 or holoTC over 6 months with planned simple contrasts, ANOVA/ANCOVA reported as F. a95% CI of ANOVA. bSignificant difference for the B12 Supplement group at P ⩽ 0.001 (group identified from planned simple contrasts). cSignificant difference for the B12 Supplement group at P ⩽ 0.025. dA Bonferonni adjustment applied, so significance accepted at P ⩽ 0.025. Dietary B12 intake analysed as change in intake over 6 months. Capsule adherence measured over 6 months.

The reported increase in dietary B12 intake for the Dietary Advice group was not associated with a significant increase in B12 biomarkers (Figure 2).

Figure 2. Mean percentage change in serum B12 and holoTC from baseline to 6 months by treatment group.

Capsule adherence and changes in B12 biomarkers Capsule adherence decreased between 2 and 6 months from 76% (95% CI (65–87)) to 59% (47–72) for the B12 Supplement group and 85% (78–93) to 59% (47–71) for the Placebo group. Capsule adherence accounted for 27% of the change in serum B12 at 2 months for the B12 Supplement group (r = 0.52, P = 0.02, (95% CI (0.11–0.78)) but a non-significant 15% of the change at 6 months (r = 0.39, P = 0.08, (−0.05, 0.7)). At 2 months, capsule adherence accounted for 22% of the change in holoTC (r = 0.47, P = 0.03, 95% CI (0.05–0.70)) but a non-significant 2% of the change at 6 months (r = − 0.04, P = 0.76, (−0.40, 0.46)). When capsule adherence was high (76%) for the B12 supplement group, it was associated with a significant change in B12 biomarkers, but when supplement adherence dropped to 59%, there was an insignificant change in B12 biomarkers. There was a nonsignificant relationship between capsule adherence and change in B12 biomarkers for the placebo group. Change in B12 intake and correlation with B12 biomarkers Reported dietary B12 intake from baseline to 6 months increased by 35% (95% CI (2–68)) in the Dietary Advice group, with no significant change in the other two treatment groups. © 2014 Macmillan Publishers Limited

DISCUSSION Half of the South Asian, Auckland, NZ-based women of childbearing age in this study had either deficient or insufficient B12 stores. Women with vegetarian dietary practices were more at risk, but non-vegetarian dietary practices did not exclude low dietary B12 intake or low B12 biomarkers. The high rate of low B12 status in this VitB12 study population is consistent with two other published studies identifying South Asian women17 and preadolescent girls28 as being more at risk from B12 deficiency due to insufficient dietary B12 intake than the general NZ population. Dietary B12 intake may also be low for other populations of women of childbearing age in NZ. In the 2008/2009 NZ Adult Nutrition Survey, 14% of all women and 22% of women aged 19–30 years reported an inadequate dietary B12 intake.29 During pregnancy, B12 deficiency may go undetected as it is not a part of the routine antenatal blood screening in NZ.30 Although adequate preconceptual folate status is justifiably emphasised,6 there is minimal emphasis on preconceptual B12 status. In 2006 (the latest published population census), South Asians numbered 3% of the NZ population of 4.2 million; double the number recorded in the 2001 census.31 These changing population demographics, plus a number of other factors, make B12 deficiency in women of childbearing age an important public health issue. These other factors include an increase in vegetarian or low meat-eating dietary practices among women,29 increased costs for B12-containing foods32 and increased consumption of medicines that inhibit B12 absorption, such as proton pump inhibitors33 and metformin.34 The VitB12 study findings support low-dose oral cyanocobalamin supplementation as being effective in improving B12 status for women with insufficient dietary B12 intake, but this improvement plateaued over 6 months. When supplement adherence was high (76%), this was associated with a significant change in B12 European Journal of Clinical Nutrition (2014) 870 – 875

Vitamin B12 deficiency in childbearing women GJ Mearns et al

874 biomarkers, but when supplement adherence decreased to 59%, B12 biomarker changes were not statistically significant. Further research is needed on appropriate supplement dose and frequency regimes. For example, a higher physiological dose (20 μg) may be more effective if taken less frequently. Research into medication adherence with bisphosphonate medicines35 and iron supplements36 found that weekly dosing schedules were associated with higher medicine adherence than daily dosing schedules. A higher dose of cyanocobalamin may accommodate a lower adherence rate and still be effective in improving B12 status. Currently, the only publicly funded medication for the treatment of B12 deficiency in NZ is a 1000 μg intramuscular B12 injection.5 A publicly funded, low-dose oral B12 supplement has potential as a non-invasive, low-cost intervention that women could selfadminister to prevent B12 deficiency. The dietary advice treatment in the VitB12 study was not associated with improved B12 biomarkers, even though median reported intake of dietary B12 increased. Although nutritional advice is recommended as a first-line strategy to address dietary deficiencies such as B12 deficiency,18 the VitB12 study findings support that more evidence is needed on how to translate that advice into effective dietary practices. Strategies such as assessing readiness for change37 and providing support groups so women can discuss ways to alter traditional foods to increase their B12 content may improve dietary B12 intake. The B12FFQ collected information on foods and quantities eaten but did not include details such as food preparation and cooking, which may influence the bioavailability of B12 from some foods such as eggs and milk.38,39 Future advice could also focus on food preparation and cooking methods that maximise B12 bioavailability. The VitB12 study was sufficiently powered to detect a significant change in B12 biomarkers in the B12 supplement group; however, it may not have been sufficient to detect a change in the Dietary Advice group. The sample size was calculated on the basis of a study with children, using milk as the supplemented food and in a controlled study situation.19 This differs from the population of childbearing age women, the recommendation of a variety of foods to be consumed and the free-living study environment of the VitB12 study. A larger number of participants may have been required in order to detect a statistically stronger and biologically meaningful difference. The lack of B12 biomarker change in the Dietary Advice group may also be because 14 out of the 20 women were already sufficient in B12 at baseline; B12 transport proteins would not upregulate to increase B12 absorption in the same way as they would during B12 deficiency.40,41 Difficulties in meeting the RDI of 2.4 μg/day42 on a vegetarian diet may also have contributed to the lack of response in the Dietary Advice group. Milk and yoghurt were the main B12 foods consumed, and the requirement for 2–3 servings per day to meet the RDI was a challenge. Findings from two studies of relationships between dietary B12 intake and B12 biomarkers suggest that the RDI may be set too low and between 4 and 7 μg dietary B12 per day is needed to maintain B12 sufficiency.43,44 Foods consumed also influence this requirement. A study in Pune, India investigated supplementation with buffalo milk.45 There was an increase in both serum B12 and holoTC concentrations and a decrease in plasma homocysteine concentrations in response to supplementation with 400 ml of buffalo milk (B12 content 2.5–3.8 μg/l) every day for 14 days for the 29 lactovegetarian participants with a serum B12 o 148 pmol/l.45 Sustainability of intake for longer periods was not investigated, and buffalo milk has a higher B12 content than that recorded for cow’s milk, but these findings are consistent with other published studies that report a high bioavailability of B12 from dairy products.43,44 This supports further research into increasing dietary intake of dairy products over a sustained period of time in order to improve B12 intake for South Asian women with lactovegetarian dietary practices. It may also be appropriate to European Journal of Clinical Nutrition (2014) 870 – 875

consider fortification of culturally preferred foods such as tofu or rice that are a staple part of the diet. B12 in fortified food (usually cyanocobalamin) does not require acid hydrolysis, hence it is better absorbed than B12 from natural B12 food sources.44 Testing for holoTC as well as serum B12 to determine B12 status provided a more comprehensive picture of B12 status for women in this study than serum B12 testing alone. HoloTC reflects cellular availability of B12 for DNA methylation and can detect tissue depletion even before the metabolites from functional deficiency start to rise.25,46 The VitB12 study used cutoff values for depletion and deficiency that reflect the risks associated with insufficient cellular B12 (Table 2)47 rather than the traditional deficiency cutoffs that reflect when macrocytic anaemia48 and neurological deficits occur.41 Although participants were stratified and randomised into groups, the groups were not evenly matched on the critical biomarker of holoTC. During deficiency, transport proteins for B12 absorption are upregulated, increasing B12 absorption.41 This may contribute to a part of the rise in holoTC for the B12 Supplement group compared with the Placebo and B12 Dietary Advice groups. There are limitations to collecting B12 dietary intake information via a FFQ with under- and over-reporting of nutritional intake well documented.49 Although the FFQ in this VitB12 study had a moderate association with B12 biomarkers, the uncertainty of B12 dietary intake throughout the treatment period could have contributed to the variance in B12 biomarkers. It was also not possible to blind the Dietary Advice group, therefore the participant and researcher knowledge of this group allocation was another potential source of bias. CONCLUSION B12 deficiency is common in South Asian women of childbearing age in Auckland, NZ, with the highest risk of deficiency in those with low or non-meat-eating dietary practices. Given the associations between maternal B12 deficiency in pregnancy and epigenetic changes that increase offspring risk of insulin resistance and neural tube defects, these findings have implications for closer monitoring and prevention of B12 deficiency in women where low or non-meat dietary practices are common. This VitB12 study provides preliminary support for physiological doses of oral cyanocobalamin supplementation as an effective and acceptable strategy for improving B12 status preconception. However, the effectiveness of the 6-μg dose in this study was associated with supplement adherence, and when adherence decreased, so did supplement efficacy. Future recommendations include researching a higher dose taken less frequently. Options also include fortification of culturally specific foods. The dietary advice treatment in this study did not improve B12 status. Further research may explicate how this advice can be delivered in a more effective, acceptable and sustainable way. CONFLICT OF INTEREST The authors declare no conflict of interest.

ACKNOWLEDGEMENTS We thank the participants in this research project and the support team: Ella Kumar (Research Assistant) and Sylvia Hobson (Phlebotomist), the research team and colleagues at the AUT Faculty of Health and Environmental Sciences. The project was funded by an AUT University Internal Contestable Research Grant and a STAR Project PhD Research Scholarship.

AUTHOR CONTRIBUTIONS GJM: study design, data collection, analysis and interpretation, drafting and final revision of the manuscript; JKM: project supervision and revision of the

© 2014 Macmillan Publishers Limited

Vitamin B12 deficiency in childbearing women GJ Mearns et al

875 manuscript, VO: data analysis and interpretation; and ECR: study design, data interpretation, project supervision and revision of the manuscript.

REFERENCES 1 Yajnik CS, Deshpande SS, Jackson AA, Refsum H, Rao S, Fisher DJ et al. Vitamin B12 and folate concentrations during pregnancy and insulin resistance in the offspring: the Pune Maternal Nutrition Study. Diabetologia 2008; 51: 29–38. 2 Bhate V, Deshpande S, Bhat D, Joshi N, Ladkat R, Watve S et al. Vitamin B12 status of pregnant Indian women and cognitive function in their 9-year-old children. Food Nutr Bull 2008; 29: 249–254. 3 Krishnaveni GV, Hill JC, Veena SR, Bhat DS, Wills AK, Karat CL et al. Low plasma vitamin B12 in pregnancy is associated with gestational 'diabesity' and later diabetes. Diabetologia 2009; 52: 2350–2358. 4 Refsum H. Folate, vitamin B12 and homocysteine in relation to birth defects and pregnancy outcome. Br J Nutr 2001; 85: S109–S113. 5 PHARMAC. Online Pharmaceutical Schedule Wellington: New Zealand Government [4 July 2013]; Available from http://www.pharmac.govt.nz/healthpros/ Schedule?code = A04 2013. 6 Food Standards Australia New Zealand Final assessment report proposal 295. Consideration of mandatory fortification with folic acid. [28 May, 2013]; Available from http://www.foodsafety.govt.nz/elibrary/industry/Proposal_P295Nzfsa_Supportive.htm 2006. 7 Food Standards Australia New Zealand Australia New Zealand Food Standards Code. Australian Government; 2013 [1 October 2013]; Available from http://www. comlaw.gov.au/Series/F2008B00615. 8 Hudson B. Vitamin B-12 deficiency. BMJ 2010; 340: c2305. 9 Samuel TM, Duggan C, Thomas T, Bosch R, Rajendran R, Virtanen SM et al. Vitamin B12 intake and status in early pregnancy among urban South Indian women. Ann Nutr Metab 2013; 62: 113–122. 10 Bhat DS, Thuse NV, Lubree HG, Joglekar CV, Naik SS, Ramdas LV et al. Increases in plasma holotranscobalamin can be used to assess vitamin B12 absorption in individuals with low plasma vitamin B12. J Nutr 2009; 139: 2119–2123. 11 Yajnik CS, Ganpule-Rao AV. The obesity-diabetes association: what is different in indians? Int J Low Extrem Wounds 2010; 9: 113–115. 12 Wadhwani NS, Pisal HR, Mehendale SS, Joshi SR. A prospective study of maternal fatty acids, micronutrients and homocysteine and their association with birth outcome. Matern Child Nutr 2013; e-pub ahead of print 25 June 2013; doi:10.1111/mcn.12062. 13 Yajnik CS, Deshpande SS, Panchanadikar AV, Naik SS, Deshpande JA, Coyaji KJ et al. Maternal total homocysteine concentration and neonatal size in India. Asia Pac J Clin Nutr 2005; 14 , 179–181. 14 Ronnenberg AG, Goldman MB, Chen D, Aitken IW, Willett WC, Selhub J et al. Preconception homocysteine and B vitamin status and birth outcomes in Chinese women. Am J Clin Nutr 2002; 76 , 1385–1391. 15 Molloy AM, Kirke PN, Troendle JF, Burke H, Sutton M, Brody LC et al. Maternal Vitamin B12 status and risk of neural tube defects in a population with high neural tube defect prevalence and no folic acid fortification. Pediatrics 2009; 123: 917–923. 16 Godbole K, Deshmukh U, Yajnik C. Nutri-genetic determinants of neural tube defects in India. Indian Pediatr 2009; 46: 467–475. 17 Gammon CS, von Hurst PR, Coad J, Kruger R, Stonehouse W. Vegetarianism, vitamin B12 status, and insulin resistance in a group of predominantly overweight/obese South Asian women. Nutrition 2012; 28: 20–24. 18 Ministry of Health. Food and Nutrition Guidelines for Healthy Adults: A Background Paper. Ministry of Health: Wellington, New Zealand, 2003 [20 September, 2013]; Available from http://www.health.govt.nz/publication/food-and-nutrition-guidelines-healthy-adults-background-paper. 19 Siekmann JH, Allen LH, Bwibo NO, Demment MW, Murphy SP, Neumann CG. Kenyan school children have multiple micronutrient deficiencies, but increased plasma vitamin B-12 is the only detectable micronutrient response to meat or milk supplementation. J Nutr 2003; 133: 3972S–3980SS. 20 Overall JE, Tonidandel S, Starbuck RR. Last-observation-carried-forward (LOCF) and tests for difference in mean rates of change in controlled repeated measurements designs with dropouts. Social Sci Res 2009; 38: 492–503. 21 Hopkins W. A New View of Stats. Sportscience: Auckland, New Zealand, 2007 [6 June, 2011]; Available from http://sportsci.org/resource/stats/index.html. 22 De Benoist B. Conclusion of a WHO technical consultation on folate and B12 deficiencies. Food Nutr Bull 2008; 29: S238–S244. 23 World Health Organization. Introduction. Folate and vitamin B12 deficiencies: proceedings of a WHO technical consultion. Food Nutr Bull 2008; 29: S3–S4. 24 Gibson RS. Principles of Nutritional Assessment, 2nd edn. Oxford University Press: New York, USA; Auckland, NZ, 2005.

© 2014 Macmillan Publishers Limited

25 Hvas AM, Nexo E. Holotranscobalamin—a first choice assay for diagnosing early vitamin B deficiency? J Intern Med 2005; 257: 289–298. 26 Field AP. Discovering Statistics Using SPSS, 3rd edn. SAGE Publications: Thousand Oaks, CA, USA, 2009. 27 Allison PD. Change scores as dependent variables in regression analysis. Sociol Methodol 1990; 20: 93–114. 28 Rush EC, Chhichhia P, Hinckson E, Nabiryo C. Dietary patterns and vitamin B12 status of migrant Indian preadolescent girls. Eur J Clin Nutr 2009; 63: 585–587. 29 University of Otago, New Zealand Ministry of Health. A Focus on Nutrition: Key Findings from the 2008/2009 New Zealand Adult Nutrition Survey. Ministry of Health: Wellington, New Zealand, 2011 [30 September 2013]; Available from http://www.health.govt.nz/publication/focus-nutrition-key-findings-2008-09-nzadult-nutrition-survey. 30 BPACNZ. Routine laboratory testing during pregnancy. BPACNZ: Dunedin, New Zealand, 2011 [30 September 2013]; Available from http://www.bpac.org.nz/ resources/bt/2011/07_pregnancy.asp#1st. 31 Statistics New Zealand. Asain Ethnic Group Profiles 2006. New Zealand Government: Wellington, New Zealand, 2006 [20 September 2013]; Available from http:// www.stats.govt.nz/browse_for_stats/people_and_communities/asian-peoples/ asian-ethnic-grp-profiles-06-tables.aspx. 32 Statistics New Zealand. Food Price Index Review. New Zealand Government: Wellington, New Zealand, 2011 [20 September 2013]; Available from http://www. stats.govt.nz/browse_for_stats/economic_indicators/prices_indexes/fpi-review2011.aspx. 33 Hirschowitz BI, Worthington J, Mohnen J. Vitamin B12 deficiency in hypersecretors during long-term acid suppression with proton pump inhibitors. Aliment Pharmacol Ther 2008; 27: 1110–1121. 34 William AB, Spencer S, Elizabeth J, Ann MS, Victor H. Increased intake of calcium reverses vitamin B12 malabsorption induced by metformin. Diabetes Care 2000; 23: 1227. 35 Silverman S, Gold D. Compliance and persistence with osteoporosis therapies. Curr Rheumatol Rep 2008; 10: 118–122. 36 Beard JL. Effectiveness and strategies of iron supplementation during pregnancy. Am J Clin Nutr 2000; 71: 1288S–1294SS. 37 Molaison EF. Stages of change in clinical nutrition practice. Nutr Clin Care 2002; 5: 251–257. 38 Watanabe F. Vitamin B12 sources and bioavailability. Exp Biol Med 2007; 232: 1266–1274. 39 Levine AS, Doscherholmen A. Vitamin B12 bioavailability from egg yolk and egg white: relationship to binding proteins. Am J Clin Nutr 1983; 38: 436–439. 40 Herrmann W, Schorr H, Purschwitz K, Rassoul F, Richter V. Total homocysteine, vitamin B12, and total antioxidant status in vegetarians. Clin Chem 2001; 47: 1094–1101. 41 Herbert V. Staging vitamin B12 (cobalamin) status in vegetarians. Am J Clin Nutr 1994; 59: 1213S–1222S. 42 The Concise New Zealand Food Composition Tables [database on the Internet] Ministry of Health 2009. Available from http://www.foodcomposition.co.nz/concise-tables. 43 Vogiatzoglou A, Smith AD, Nurk E, Berstad P, Drevon CA, Ueland PM et al. Dietary sources of vitamin B-12 and their association with plasma vitamin B-12 concentrations in the general population: the Hordaland Homocysteine Study. Am J Clin Nutr 2009; 89: 1078–1087. 44 Tucker KL, Rich S, Rosenberg I, Jacques P, Dallal G, Wilson PW et al. Plasma vitamin B-12 concentrations relate to intake source in the Framingham Offspring Study. Am J Clin Nutr 2000; 71: 514–522. 45 Naik S, Bhide V, Babhulkar A, Mahalle N, Parab S, Thakre R et al. Daily milk intake improves vitamin B-12 status in young vegetarian Indians: an intervention trial. Nutr J 2013; 12: 136. 46 Herrmann W, Obeid R, Schorr H, Geisel J. The usefulness of holotranscobalamin in predicting vitamin B12 status in different clinical settings. Curr Drug Metab 2005; 6: 47–53. 47 Herrmann W, Obeid R, Schorr H, Geisel J. Functional vitamin B12 deficiency and determination of holotranscobalamin in populations at risk. Clin Chem Lab Med 2003; 41: 1478–1488. 48 Oosterhuis WP. Diagnostic value of the mean corpuscular volume in the detection of vitamin B12 deficiency. Scand J Clin Invest 2000; 60: 9–18. 49 Kipnis V, Subar AF, Midthune D, Freedman LS, Ballard-Barbash R, Troiano RP et al. Structure of dietary measurement error: results of the OPEN biomarker study. Am J Epidemiol 2003; 158: 14–26. 50 Schulz KF, Altman DG, Moher D. CONSORT 2010 statement: Updated guidelines for reporting parallel group randomised trials. J Pharmacol Pharmacother 2010; 1: 100–107.

European Journal of Clinical Nutrition (2014) 870 – 875