Increases in Plasma Holotranscobalamin Can Be Used to Assess Vitamin B-12 Absorption in Individuals with Low Plasma Vitamin B-12^,,

Abstract

Low plasma concentrations of vitamin B-12 are common in Indians, possibly due to low dietary intakes of animal-source foods. Weather malabosrption of the vitamin contributes to this has not been investigated. A rise in plasma holotranscobalamin (holo-TC) concentration after a standard dose of oral vitamin B-12 has been proposed as a measure of gastrointestinal absorption in people with normal plasma vitamin B-12 concentrations. We studied 313 individuals (children and parents, 109 families) in the Pune Maternal Nutrition Study. They received 3 doses of 10 μg (n=191) or 2 μg (n=122) of cyanocobalamin at 6 h intervals. A rise in plasma holo-TC of ≥15% and >15 pmol/L above baseline was considered normal vitamin B-12 absorption. The baseline plasma vitamin B-12 was <150 pmol/L in 48% of participants; holo-TC was <35 pmol/L in 98%, and total homocysteine was high in 50% (>10 μmol/L in children and >15 μmol/L in adults). In 10 μg group plasma holo-TC concentration increased by 4.8 -fold from (mean ± SD) 9.3 ± 7.0 pmol/L to 53.8 ± 25.9 pmol/L, and in 2 μg group by 2.2 -fold from 11.1 ± 8.5 pmol/L to 35.7 ± 19.3 pmol/L. Only 10% of participants, mostly fathers, had an increase less than the suggested cut-points. Our results suggest that an increase in plasma holo-TC may be used to assess vitamin B-12 absorption in individuals with low vitamin B-12 status. Because malabsorption is unlikely to be a major reason for the low plasma vitamin B-12 concentrations in this population, increasing dietary vitamin B-12 should improve their status.

Introduction

Low vitamin B-12 concentrations are common in Indians (1) and contribute to hyperhomocysteinemia despite normal folate status (2, 3). This is unlike the situation in Europeans (4), where low folate status is the predominant determinant of hyperhomocysteinemia. We have demonstrated an association between low maternal vitamin B-12 concentrations in pregnancy and insulin resistance (5) and reduced neurocognitive performance in the children (6). Vitamin B-12 deficiency could also contribute to the etiology of neural tube defects in Indians (7). Prevention of these conditions in India may require supplementation with vitamin B-12.

Low vitamin B-12 status among Indians is usually attributed to low dietary intake due to a mainly vegetarian diet (1, 8, 9) but it is not uncommon even in those who eat non-vegetarian foods (9). In our studies, even though vegetarian diet was a significant predictor of low vitamin B-12 status, one half to two thirds of participants who eat non-vegetarian food more than 3 d/wk also had low plasma vitamin B-12 concentrations (C. S. Yajnik, D. S. Bhat, H. G. Lubree, C. V. Jojalekar, unpublished data). We also found an inverse association of plasma vitamin B-12 concentrations with levels of education, income and hygiene, suggesting that factors other than low dietary content could contribute to vitamin B-12 status (3). Tropical sprue, and gastrointestinal infestations (10, 11) and infections (including Helicobacter pylori) (12, 13) also contribute to low vitamin B-12 concentrations, possibly due to malabsorption. This possibility is rarely investigated because of the complexity of methods for investigating vitamin B-12 absorption. For example, the Schilling (14) test requires radioactive isotope. It is essential to determine whether the low vitamin B-12 status among Indians only reflects dietary inadequacy or if malabsorption is also a contributing factor, becasuse this will have implications for treatment.

Recently, Bor et al (15) demonstrated an increase in plasma holotranscobalamin (holo-TC) concentration after a small oral dose of vitamin B-12 (9 μg × 3 doses), and suggested that this approach can be used to test vitamin B-12 absorption in a clinical setting. Another recent study by van Castel-Roberts et al (16) investigated circulating concentrations of holo-TC using Bor’s protocol in a sample of healthy US men and women. Blood samples taken 24 h after the first dose provided the optimal results. Both these studies were performed in participants with normal vitamin B-12 concentrations. We have used the same protocol with slight modification (10 μg × 3 doses) as a test of vitamin B-12 absorption in a population with low vitamin B-12 status. In addition, we also investigated the response, using a much lower dose (2 μg × 3 doses).

Participants and methods

The participants were families from an extended cohort of the Pune Maternal Nutrition Study (PMNS). The design and methods have been reported elsewhere (17). In short, the PMNS is a prospective community-based study to investigate the relationship of maternal nutrition to fetal growth; the participants have been followed up to assess future risk of type 2 diabetes and cardiovascular disease. The study was established in 6 villages 40-50 km from Pune city, between June 1994 and April 1996 and covered a population of >35,000. We screened all married non-pregnant women who were < 35 y and not sterilized, were screened. Of 2,675 eligible women, 2,466 (92%) agreed to participate. During the study period 814 women became pregnant and participated in the main PMNS cohort. After the main study, we enrolled an additional 153 pregnant women from the same initial sample of eligible women, to study early fetal growth. Of these, 129 delivered in the study and the babies and parents have been followed up with the original cohort; 119 children remained in follow-up at 9 y of age. These children and their parents were invited to take part in the current study, which took place from May–November 2006. The women, their husbands and the children, were comparable to the main PMNS cohort in body size measurements and socioeconomic status. Children as well as adults were included in the study because our earlier work suggested that vitamin B-12 status varies with age.

The King Edward Memorial (KEM) Hospital Ethics Committee approved the study, and we obtained informed written consent from the parents, and assent from the children.

The participants reported to the Diabetes Unit, KEM Hospital, Pune, the evening before the study. Height was measured to the nearest 0.1 cm using a wall-mounted stadiometer (CMS Instruments), and body weight to the nearest 5 g (Conveigh, Electronic Instruments). A standard vegetarian dinner was provided between 20-21 h, after which they rested overnight in the Diabetes Unit. The only food of animal origin during period was a small amount of milk in tea. All participants were examined for gross clinical signs of protein-energy under-nutrition and vitamin deficiencies (vitamins A, B complex, C, D). In children, a FFQ was administered to record the portion size and frequency of consumption of different foods, over a period of 1 y. Vitamin B-12 intakes were calculated from the FFQ data using published tables of the nutrient content of Indian foods (18). A fasting blood sample was collected in the morning (baseline), and a 10 μg vitamin B-12 (cyanocobalamin) capsule was administered under supervision every 6 h thereafter (8, 14 and 20 h). A fasting blood sample was collected the following morning, ~ 12 h after the last dose (post-dose sample).

After completing sampling for approximately one-half of the families, it was clear from our data that absorption was satisfactory in most cases. We then reasoned that we could improve the utility of the study, including for policy makers, by investigating a smaller dose, nearer to the Recommended Dietary Allowance (12) of 2.4 to 2.8 μg for adults and 0.9 to 2.4 μg for children. The remaining families received 2 μg capsules, rather than 10 μg capsules. In total, 65 families received 10 μg × 3 doses (10 μg group) and 44 received 2 μg × 3 doses (2 μg group ).

Laboratory analysis

Blood samples from the ante-cubital vein were collected in the sitting position, in an EDTA vacutainer. Hemoglobin was measured on a Beckman Coulter Analyzer (A^C.T diff^TMAnalyzer). The remaining blood was centrifuged at 2500 x g for 15 min at 4°C within 1h of collection, and plasma was stored at −70°C until further analysis. Plasma cobalamin (B-12) was measured by microbiological assay using a colistin sulfate-resistant strain of Lactobalillus. leichmanii (19, 20). Plasma holo-TC was measured using magnetic beads (microspheres) with immobilized monoclonal antibody specific for human TC (21) followed by the conventional microbiological assay developed for cobalamin estimation (22). Plasma folate was measured by microbiological assay using a chloramphenicol-resistant strain of Lactobalillus. casei (23, 24). Plasma total homocysteine (tHcy) was measured by fluorescence polarization immunoassay (Abbott) (25). In our laboratory, for vitamin B-12, holo-TC, folate, and tHcy analysis, between-day CV were < 8%, <9%, <7%, and <3%, respectively.

Definitions

We defined anemia as a hemoglobin concentration below 120 g/L in children and mothers, and below 130 g/L in fathers (26). We defined microcytosis as mean corpuscular volume as <80 fL and macrocytosis and >100 fL. Hyperhomocysteinemia was defined as plasma tHcy concentration >15 μmol/L in adults, and >10 μmol/L in children (27). Low folate concentrations were defined as plasma folate concentration <7.0 nmol/L (28). Low vitamin B-12 concentrations were defined as plasma vitamin B-12 concentration <150 pmol/L (2). Low holo-TC concentration was defined as <35 pmol/L (29). Poor vitamin B-12 absorption was defined by the Bor et al. (16) criteria as a rise in plasma holo-TC <15%, and <15 pmol/L after 3 doses of 10 μg oral vitamin B-12.

Statistical methods

The data are presented as mean ± SD. Differences between baseline and post-dose measurements (vitamin B-12, tHcy, and holo-TC) were tested using paired t-tests. Differences between groups were tested using unpaired t-tests. Results in children and parents are analyzed separately. The results in children and parents for 2 μg and 10 μg groups were combined in the multiple linear regression analysis of factors associated with rise in plasma holo-TC. SPSS version 11.0 for windows was used for statistical analysis.

Results

A total of 313 individuals (109 children, 96 fathers, and 108 mothers) participated in the study. None were taking vitamin supplements or drugs known to influence vitamin B-12 absorption such as proton pump inhibitors and metformin. One child from 10 μg group could not swallow the capsules, and one child from 2 μg group vomited during the test; these were excluded from the analysis.

None of the children, fathers, and mothers had clinical signs of protein-energy under-nutrition or vitamin deficiencies; some were anemic (Table 1). Sixty four percent of anemic participants had microcytic erythrocytes, none had macrocytosis. Five participants (3 fathers and 2 mothers) had macrocytic erythrocytes but were not anemic. FFQ data were available for children only. None of them were vegan. One-third were lacto-vegetarian (milk but no other non-vegetarian foods) and two-thirds ate non-vegetarian foods (eggs, meat, and fish as well as milk). Less than one-half of the later group (44%) consumed non-vegetarian foods more than twice a week, but the average portion size of non-vegetarian foods was small. The median calculated daily dietary intake of vitamin B-12 was <0.1 μg in the lacto-vegetarian children and 0.2 μg in the non-vegetarian children.

TABLE 1.

Anthropometric characteristics and hemoglobin concentrations of children, fathers and mothers¹.

	Children	Fathers	Mothers
n	109	96	108
Age, y	9.0 ± 0.2	36.9 ± 3.8	30.2 ± 3.0
Height, cm	126.5 ± 5.6	165.7 ± 7.1	153.2 ± 5.3
Weight, kg	21.9 ± 2.9	59.3 ± 10.0	47.7 ± 8.2
Hemoglobin, g/L	125.0 ± 1.0	143.0 ± 1.4	120.0 ± 1.6
Anemic2, %	24	13	46

¹Values are mean ±SD, unless specified

²Hemoglobin <120 g/L for children, and mother, and <130 g/L for father (26)

Baseline plasma vitamin B-12 concentrations were low, folate concentrations were normal, and tHcy concentrations were high (Table 2). Twenty seven percent of children, 70% of fathers, and 49% of mothers had low vitamin B-12 concentrations. Ninety eight percent of participants had low plasma holo-TC concentrations. In contrast, only 2% of children, 15% of fathers, and 9% of mothers had low folate concentrations. Hyperhomocysteinemia was observed in 48% of children, 74% of fathers, and 34% of mothers.

TABLE 2.

Plasma vitamin B-12, holo-TC, tHcy and folate concentrations in Indian children and their parents before and after receiving 3 × 10 μg doses or 3 × 2 μg doses of cyanocobalamin¹.

	10 μg group			2 μg group
	Children	Fathers	Mothers	Children	Fathers	Mothers
n	64	61	65	43	35	43
Plasma vitamin B-12, pmol/L
Baseline	191 ± 60	136 ± 62	180 ± 176	222 ± 107	126 ± 68	161 ± 95
Post-dose	302 ± 95	193 ± 74	233 ± 121	293 ± 123	158 ± 71	194 ±103
Change %	63.1 ± 39.6*	47.7 ± 31.1*	40.1 ± 24.0*	38.9 ± 27.4*	31.3 ± 21.1*	23.5 ± 14.9*
Plasma Holo-TC2, pmol/L
Baseline	10.3 ± 5.6	7.7 ± 4.2	9.7 ± 9.7	11.9 ± 7.3	10.5 ± 8.1	10.7 ± 10.0
Post-dose	67.4 ± 29.5	45.7 ± 22.2	47.9 ± 19.4	49.5 ± 16.3	24.9 ± 15.7	30.5 ± 17.0
Change %	655.6 ± 406.8*	577.5 ± 406.3*	498.6 ± 280.5*	399.6 ± 207.0*	150.6 ± 99.9*	229.2 ± 143.1*
Plasma tHcy, μmol/L
Baseline	11.0 ± 3.7	29.2 ± 19.2	15.3 ± 8.3	10.3 ± 4.0	35.9 ± 27.5	14.1 ± 7.1
Post-dose	10.0 ± 3.4	27.8 ± 18.2	14.5 ± 7.7	9.6 ± 3.9	32.9 ± 25.7	12.1 ± 6.1
Change %	−8.2 ± 12.5*	−4.1 ± 11.8*	−3.7 ± 21.6*	−6.1 ± 11.8*	−8.2 ± 11.5*	−12.5 ± 16.3*
Plasma folate, nmol/L
Baseline	15.9 ± 5.0	10.4 ± 3.7	13.9 ± 7.0	15.7 ± 4.5	11.1 ± 4.2	12.8 ± 7.1
Post-dose	16.4 ± 5.5	10.6 ± 3.6	13.0 ± 6.9	14.8 ± 4.2	10.9 ± 4.2	11.5 ± 6.9
Change %	9.8 ± 42.5	3.4 ± 17.0	−6.2 ± 20.3*	−3.0 ± 24.5	−1.1 ± 16.6	−8.6 ± 20.6*

¹Values are mean ±SD,

^*Significant change. P< 0.05.

²Individual data in supplemental Figure 1

Plasma holo-TC concentrations increased after oral cynocobalamin (Table 2, Supplemental Figure 1). In 10 μg group the rise was 4.8-fold and in 2 μg group the rise was 2.2-fold. Plasma vitamin B-12 concentrations also increased, and the two were related (r=0.57; P<0.001, adjusted for age and sex). The increases in holo-TC and vitamin B-12 were greater in children than in the parents (P<0.001), and comparable in fathers and mothers. This may have resulted from a greater dose of vitamin B-12 per kg body weight in children (10 μg group: 1.39 ± 0.17 μg/kg; 2 μg group: 0.28 ± 0.03 μg/kg) compared with fathers (10 μg group: 0.53 ± 0.09; 2 μg group: 0.10 ± 0.02) and mothers (10 μg group: 0.65 ± 0.10; 2 μg group: 0.13 ± 0.02).

According to the criteria listed by Bor et al (15) (used only for 10 μg group), four children (6%), 10 fathers (17%), and 5 mothers (8%) were classified as poor vitamin B-12 absorbers. The poor absorbers and normally absorbing counterparts had similar baseline plasma vitamin B-12, holo-TC, tHcy and blood hemoglobin concentrations; none of poor absorbers had macrocytic erythrocytes.

In a multivariate analysis, combining children and parents from both the groups, and including age, sex, dose/kg of cyanocobalamin, and baseline vitamin B-12 concentration, the rise in plasma holo-TC concentration was directly proportional to the dose/kg of cyanocobalamin (standardized β=0.446; P<0.001) and inversely related to age (standardized β= −0.147; P=0.01). The rise in holo-TC was higher in participants with a baseline plasma vitamin B-12 concentration >150 pmol/L at baseline than in those with concentrations <150 pmol/L (unadjusted (mean ± SEM) 42.3 ± 1.8 pmol/L vs 30.6 ± 1.9 pmol/L; p=0.005; and 42.0 ± 1.6 pmol/L vs 30.9 ± 1.6 pmol/L, p=0.06, adjusted for age, sex, dose/kg of vitamin B-12 and body weight).

After oral cyanocobalamin, plasma folate concentrations did not change, but those of tHcy decreased from 18.3 ± 14.4 to 17.1 ± 13.5 μmol/L in 10 μg group and from 19.0 ± 18.9 to 17.0 ± 17.5 μmol/L in 2 μg group (p<0.001).

Discussion

We have demonstrated the usefulness of the Bor et al. (15) protocol to investigate gastrointestinal absorption of vitamin B-12 in participants with low circulating vitamin B-12 concentrations. The absolute rise in holo-TC in our participants (44 pmol/L) was comparable to that in vitamin B-12-sufficient Europeans (46 pmol/L) (15); however, the percentage rise was much higher (~550% compared with ~50%) because of lower baseline concentrations (median 7 pmol/L compared with 72 pmol/L). Our findings provide experimental evidence that vitamin B-12 absorption is adequate in the majority of Indians and argue against malabsorption as a prominent cause of their low vitamin B-12 concentrations. Furthermore, it was reassuring that a physiological dose of cyanocobalamin (2 μg × 3) was absorbed satisfactorily, suggesting that a modest increase in vitamin B-12 intake should improve their status.

Vitamin B-12 in food is protein-bound. It is released in the stomach by the action of pepsin and gastric acid, and then combines with salivary R-binder. This complex is broken down by pancreatic proteases in the duodenum, to release free B-12, which combines with intrinsic factor, secreted by gastric parietal cells (30). The intrinsic factor-vitamin B-12 complex interacts with a specific receptor in terminal ileal enterocytes, and is absorbed by endocytosis (31, 32). During transport across the enterocyte, vitamin B-12 is complexed with transcobalamin to form holo-TC, which is released into the blood (31). Unlike protein-bound vitamin B-12 in food (33), the free vitamin B-12 (cyanocobalamin) used in our study binds directly to R-binder, bypassing the stage of gastric degradation. From this stage onwards, vitamin B-12 from food and free B-12 follow the same absorption pathway. Thus, our study does not test gastric degradation but suggests normal handling beyond this stage. Holo-TC has a short half-life (34, 35) and is therefore proposed to reflect recent vitamin B-12 absorption. However, circulating holo-TC concentrations are also influenced by hepatic and renal uptake, production, and release from the ileum and kidney, tissue requirements, and other unknown factors (36). There is a debate about whether basal holo-TC concentrations represent ‘an early general cobalamin insufficiency or specifically decreased cobalamin absorption’ (36). A dose-dependent rise in holo-TC during the 24 hours following vitamin B-12 administration, suggests that the levels are influenced by recent absorption.

The rise in holo-TC was ~ 28% lower (P=0.005) in those with low vitamin B-12 concentrations compared to that in those with normal vitamin B-12 concentration. This raises the possibility that malabsorption may play a contributory role in the low vitamin B-12 concentrations in this population. It would be interesting to measure the rise in holo-TC before and after vitamin B-12 repletion, to confirm whether adequate vitamin B-12 status changes vitamin B-12 absorption.

Fewer than 10% of the participants in our study had a lower than expected rise in holo-TC based on the Bor et al. protocol (15). This may be evidence of poor vitamin B-12 absorption, in these individuals, but because the Bor protocol was developed in populations with normal vitamin B-12 status, weather our findings in some participants were due to malabsorption needs further investigation (37). However, the original test using a dose of 9 μg × 3 has been validated in inherited syndromes of vitamin B-12 malabsorption (38). In our population, more parents (12%) than children (6%) had a lower than expected rise in holo-TC, which may be due to the relatively lower dose of vitamin B-12 per kilogram body weight they received. Vitamin B-12 absorption also decreases with age (39, 40), and deficiency is common in the elderly, even in Western non-vegetarian populations (41). One of the proposed mechanisms is an increasing prevalence of ‘chronic gastritis’ with age, which results in hypochlorhydria and impaired synthesis of intrinsic factor and pepsin. The etiology of chronic gastritis includes autoimmune damage (pernicious anemia) and H. pylori infection (12, 13, 31). Further studies are required to identify the cause(s) of the low plasma vitamin B-12 in these participants.

Our study has many strengths. It was community-based, had a large sample-size, high participation rates, and low drop-out rates. We enrolled children as well as adults, and included both genders. None of the participants were taking vitamin supplements, and care was taken to ensure no interference from diet during the test period by requiring that vegetarian diets be consumed. Participants were admitted to our research unit to ensure that the test was carried out under strict supervision. In addition to the diagnostic dose we also studied absorption of vitamin B-12 at a dose close to the physiological level, which may be important in guiding public health policy. Our study cannot be strictly compared to the Bor et al (15), because we used 10 μg × 3 doses instead of 9 μg × 3 doses because of availability. We do not think that the slightly higher dose is likely to change our conclusions because absorption was similarly impressive for the 2 μg × 3 doses. A definite limitation of the study is that because we used free cyanocobalamin, we could not test the gastric degradation step which is essential for absorption of protein-bound dietary vitamin B-12 (33). However, the test would probably be able to detect the malabsorption related to tropical sprue or gastrointestinal infections, which are thought to contribute to subclinical vitamin B-12 status in tropical areas. The holo-TC test is simple, relatively inexpensive, and has the potential to be used in population-based studies, but further work regarding dosing, and validation against a gold standard test, would be desirable.

Thus, we have demonstrated adequate intestinal absorption of free cyanocobalamin in a community-based study in a South Asian Indian population with high prevalence of low vitamin B-12 concentrations. The health implications of low vitamin B-12 concentrations are not fully understood, because there is poor association with macrocytosis or severe anemia, or neurological symptoms, and also the fact that this population has perhaps lived with such low concentrations for generations. This issue needs to be further investigated. This study is a small part of such a systematic programme of research. Our results suggest that the gastro-intestinal malabsorption of the vitamin is unlikely to be a prominent cause of the high prevalence of low vitamin B-12 concentrations, and the consequent hyperhomocysteinemia in children and young adults, though this could be a contributing factor in middle-aged and elderly adults. The marked increase in plasma concentrations following even low doses of vitamin B-12 suggests that fortification of food items or regular use of supplements would be useful approaches. The results of such long-term supplementation in this population are described in a separate paper.

ABBREVATIONS USED

holo-TC

holotranscobalamin

KEM

King Edward Memorial

PMNS

Pune Maternal Nutrition Study

tHcy

total homocysteine