A dual-isotope-labeling method of studying the bioavailability of hexaglutamyl folic acid relative to that of monoglutamyl folic acid in humans by using multi

¹ From the Wageningen Centre for Food Sciences, Wageningen, Netherlands (AM-B, PV, and MBK); the Division of Human Nutrition, Wageningen University, Wageningen, Netherlands (AM-B, PV, CEW, MBK, and FJK); the State Institute for Quality Control of Agricultural Products, Wageningen, Netherlands (JAvR and JJPL); and the Department of Medicinal Chemistry and Pharmacognosy, University of Illinois College of Pharmacy, Champaign, IL (RBvB and SDG)

² Clive West died on 27 August 2004.

³ Supported by the Wageningen Centre for Food Sciences, an alliance of major Dutch food industries, Maastricht University, TNO Quality of Life, Wageningen University and Research Centre, the Dutch government, and grant no. R03 CA10331 from the National Cancer Institute.

⁴ Reprints not available. Address correspondence to P Verhoef, Department of Nutrition, Unilever Food and Health Research Institute, Olivier van Noortlaan 120, 3130 AC Vlaardingen, Netherlands. E-mail: petra.verhoef{at}unilever.com.

ABSTRACT  
Background: The bioavailability of dietary folate may be hampered by the need of the glutamate moieties to be deconjugated before absorption. Previous studies comparing the bioavailabilities of polyglutamyl and monoglutamyl folic acid had inconsistent results.

Objective: The objective was to estimate the bioavailability of polyglutamyl relative to that of monoglutamyl folic acid by using a sensitive stable-isotope approach that allowed for the administration of multiple low doses in humans.

Design: Twenty subjects aged 20–50 y ingested 2 capsules daily for 28 d; each capsule contained 50 nmol [¹³C₆]hexaglutamyl and 50 nmol [¹³C₁₁]monoglutamyl folic acid. Amounts of the isotopically labeled compounds in the capsules were verified by various methods. The degrees of isotopic enrichment of plasma 5-methyltetrahydrofolate with ¹³C₆ and ¹³C₁₁ were measured by using liquid chromatography tandem mass spectrometry, and the ratio of ¹³C₆ to ¹³C₁₁ (¹³C₆:¹³C₁₁) in plasma on day 28 was used as a measure of their relative bioavailability.

Results: The ¹³C₁₁:¹³C₆ in plasma 5-methyltetrahydrofolate reached equilibrium on day 4 and was 0.66 (95% CI: 0.58, 0.74) on day 28. The ¹³C₁₁:¹³C₆ content in the capsules varied between 1.18 and 1.96. After correction for this ratio, the estimated bioavailability of hexaglutamyl relative to that of monoglutamyl folic acid was 78%.

Conclusion: Multiple dosing of low amounts of labeled folic acid is a sensitive, accurate, and efficient method of measuring the relative bioavailability of folic acid compounds, provided that the administered doses can be reliably assessed.

Key Words: Folic acid • bioavailability • stable isotopes • humans • mass spectrometry

INTRODUCTION  
Folate deficiency in humans leads to anemia (1), neural tube defects (2), and, possibly, chronic diseases such as cardiovascular disease (3), colon cancer (4), and neurocognitive dysfunction (5). Folate status is ascertained not only by the intake of folate but also by its bioavailability. Bioavailability is defined as the proportion of ingested folate that is absorbed and available for metabolic processes and storage (6). In humans, the bioavailability of folate from the diet is assumed to be 50% (7), whereas the bioavailability of synthetic folic acid used in supplements and as a food fortificant is estimated to range from 76% to 97% (8).

Folate from the diet is to a large extent conjugated to a polyglutamate chain (9-11). Before absorption, the polyglutamate chain is deconjugated by the enzyme folylpoly -glutamate carboxypeptidase in the jejunum of the small intestine. Folate is subsequently absorbed and transported in the body as a monoglutamate. A limited capacity for deconjugation may be a cause of folate's lower bioavailability from foods (12).

Estimates of the bioavailability of polyglutamyl folic acid relative to that of monoglutamyl folic acid in humans vary widely between persons and between studies (13). These variances probably reflect differences in the dosing, timing, and route of administration of the folate compounds (14-25). Folate from the diet is usually ingested in small quantities spread over the day. It may well be that small doses of polyglutamate folate—eg, 25–50 µg—can be readily deconjugated by folylpoly -glutamate carboxypeptidase, whereas the capacity for deconjugation is insufficient for higher single doses.

To address these differences in the capacity for deconjugation, we developed a method in which specifically ¹³C-labeled monoglutamyl folic acid and specifically ¹³C-labeled polyglutamyl folic acid would be administered in multiple low doses over a certain period. We assumed that the folate carrying the specific tags from each of the 2 labeled compounds would mix completely with folate in the body pools. Hence, we anticipated that enrichment—ie, the concentration of labeled folate relative to that of unlabeled folate—would reach a plateau over time. The ratio of [¹³C₆] to [¹³C₁₁] folate ([¹³C₆:¹³C₁₁] folate) in the plasma, corrected for the ratio of the doses of the 2 labeled compounds administered, would provide a measure of the relative bioavailabilities of polyglutamyl and monoglutamyl folic acids (26). The time required to reach a plateau in the degree of enrichment and the between-person variation in the estimate of bioavailabilty were examined. In addition, we compared the performance of 2 liquid chromatography tandem mass spectrometry (LC-MS/MS) methods used for measuring the degree of isotopic enrichment.

SUBJECTS AND METHODS  
The first steps in the development of the methods were to synthesize the specifically ¹³C-labeled monoglutamate and polyglutamate folic acids and to develop LC-MS/MS methods to measure the anticipated low plasma concentrations of isotopic enrichment in folate.

Labeled compounds
[¹³C₁₁]Monoglutamyl folic acid and [¹³C₆]hexaglutamyl folic acid were synthesized specifically for our studies by ARC (Apeldoorn, Netherlands). The ¹³C atoms were incorporated into the p-aminobenzoic acid structure (6 ¹³C atoms) and, for the monoglutamyl variant, also in the glutamyl moiety (5 additional ¹³C atoms). We incorporated at least six ¹³C atoms to offset the mass of the labeled folate from that of the natural ¹³C isotopologues of the unlabeled compound. The procedure was based on the method of Maunder et al (27). The chemical purity of the labeled folic acid was >96% as measured by using nuclear magnetic resonance, optical rotation, and element counting, and the isotopic incorporation was >98% as measured by mass spectrometry.

Liquid chromatography tandem mass spectrometry procedure
We aimed to develop an LC-MS/MS method that would be sensitive enough to measure concentrations of isotopically labeled folate at 1% of the normal concentration in plasma. Given that the average plasma folate concentration is 10 nmol/L, the limit of quantification (LOQ) of the method should be 100 pmol/L. For this purpose, an LC-MS/MS method using positive ionization mode was developed that could measure [¹³C₆] and [¹³C₁₁]5-methyltetrahydrofolate (5-MTHF) in the presence of abundant unlabeled [¹³C₀]5-MTHF in plasma in the required ranges [State Institute for Quality Control of Agricultural Products (RIKILT), Wageningen, Netherlands). This method is referred to as the Wageningen LC-MS/MS method.

Sample preparation
Plasma (500 µL) was mixed with 1 mL buffer (1% wt:vol ascorbate, 100 mmol ammonium acetate/L). Sample preparation was carried out by using solid-phase extraction on an Oasis MAX cartridge (Waters, Milford, MA). The column was conditioned with 1 mL methanol and then with 1 mL buffer (1% wt:vol ascorbic acid, 100 mmol ammonium acetate/L). Diluted plasma samples were loaded on the solid-phase extraction cartridge, which was rinsed with 2 mL 50% vol:vol methanol, vacuum dried, and rinsed again with 2 mL ethyl acetate. After a second vacuum drying, the solid-phase extraction cartridge was eluted with 2 mL of 3% vol:vol formic acid in methanol. The methanol fraction was dried at 50 °C under a stream of nitrogen and reconstituted in 200 µL buffer (1% wt:vol ascorbate, 100 mmol ammonium acetate/L). Aliquots of 25 µL were used for LC-MS/MS analysis.

Liquid chromatography
HPLC separation of differentially labeled 5-MTHF was performed on a reversed-phase LUNA C18 column (150 x 2.0-mm internal diameter; 10 µm film thickness; Phenomenex, Torrance, CA). An acetonitrile–5 mmol/L formic acid eluent was used with a 10-min linear gradient from 0% to 80% (by vol) acetonitrile, starting 2 min after injection; the flow rate was 250 µL/min.

Mass spectrometry
The HPLC system was directly coupled to a Quattro Ultima triple quadrupole mass spectrometer (Micromass, Manchester, United Kingdom) equipped with an electrospray interface by using the positive ionization mode. The mass spectrometer was operated with ion source settings optimized for maximum sensitivity. [¹³C₅]5-MTHF, labeled in the glutamyl moiety (Eprova, Schaffhausen, Switzerland), was used as an internal standard. Data acquisition was done by multiple reaction monitoring of the major fragmentation reaction characteristics. The multiple reaction monitoring transitions selected for quantitative LC-MS/MS analysis were as follows: mass-to-charge ratio (m/z) 466 319 for [¹³C₆]5-MTHF, m/z 471 319 for [¹³C₁₁]5-MTHF, m/z 460 313 for [¹³C₀]5-MTHF (endogenous, nonlabeled folate), and m/z 465 313 for the internal standard [¹³C₅]5-MTHF.

Quality control
The limit of detection for this method was 20–40 pmol/L, and the LOQ was 60–80 pmol/L. Recoveries were between 65% and 80%. From the measured plasma concentrations of [¹³C₁₁] and [¹³C₆]5-MTHF, the ratio of [¹³C₆]5-MTHF to [¹³C₁₁]5-MTHF ([¹³C₆:¹³C₁₁]5-MTHF) was derived for each subject at each time point. The within-run CV of these ratios was <5%.

As an independent analysis, a different LC-MS/MS method that used the negative ionization mode was developed for comparison of results (University of Illinois College of Pharmacy, Chicago, IL). This method is referred to as the Chicago LC-MS/MS method. Major differences between the Chicago and Wageningen LC-MS/MS methods were that the Chicago method included the precipitation of plasma proteins by the addition of ammonium acetate in acetonitrile during sample preparation, used a minibore hydrophilic interaction column (HILIC; The Nest Group, Southborough, MA) for HPLC separation (28), and used a TSQ Quantrum triple quadrupole mass spectrometer (Thermo Electron, San Jose, CA) in the negative ionization mode for multiple reaction monitoring acquisition.

We decided to limit measurements to 5-MTHF because this is the predominant folate compound in plasma, and concentrations of other labeled folate compounds were assumed to be below the LOQ. Several pilot studies were carried out to ensure that both LC-MS/MS methods could measure 5-MTHF enrichments in plasma in the required ranges.

Subjects
Subjects were recruited from among the staff and students of Wageningen University by general E-mail messages and poster advertisements. A total of 20 healthy subjects aged 18–50 y were included in the study. Exclusion criteria included chronic diseases such as renal insufficiency or liver or gastrointestinal disease, the use of drugs that interfere with folate metabolism, and low vitamin B-12 status (<130 pmol/L).

Subjects were informed about the design and purpose of the study both in writing and orally, and they gave written informed consent. The study protocol was approved by the Medical Ethics Committee of Wageningen University (Wageningen, Netherlands).

Capsules
Both labeled folic acid compounds were encapsulated together at target amounts of 50 nmol (23 µg) [¹³C₁₁]monoglutamyl folic acid and 50 nmol (55 µg) [¹³C₆]hexaglutamyl folic acid per capsule (Pharmacy of the Gelderse Vallei Hospital, Ede, Netherlands). The amounts of hexaglutamyl and monoglutamyl folic acid in the capsules were examined by analyzing 5 capsule samples (3 capsules/sample) that were dissolved in Ches-HEPES buffer (pH 7.8), by HPLC, a method that is referred to as the Wageningen HPLC method (RIKILT; 29). For this analysis, folic acid content was measured before and after the deconjugation of hexaglutamyl folic acid to the monoglutamyl form, which showed monoglutamyl and total folic acid contents, respectively. Deconjugation was carried out by using rat plasma, and the incubation time was 4 h at 37 °C. Completeness of deconjugation was confirmed by the addition of triglutamyl folic acid. Hexaglutamyl folic acid content was subsequently calculated as the difference between total folic acid and monoglutamyl folic acid contents. A quality-control sample was included in the analyses, and values were within 10% of the target values. Calibration lines were checked by using molar absorption coefficients of the vitamers (29). According to these HPLC analyses, the amount of monoglutamyl folic acid in the capsules was 12% lower than expected, and the amount of hexaglutamyl folic acid was 36% lower. These differences from expected amounts resulted in a monoglutamyl-to-polyglutamyl folic acid ratio of 1.37 in the capsules.

Because it was unclear to us why the amounts of monoglutamyl and especially of polyglutamyl folic acid in the capsules were lower than expected, we decided to verify the results by other methods. The classical microbiological method using Lactobacillus casei (TNO Food and Nutrition Research, Zeist, Netherlands) proved not to be reliable in this case because the CV between duplicate samples was high (13–19%). If the deconjugation of hexaglutamyl folic acid were incomplete (despite the triglutamyl control), we could have underestimated the actual amount of hexaglutamyl folic acid in the capsules. Therefore, another HPLC method (the Florida HLPC method; University of Florida, Gainesville, FL) that did not require prior deconjugation was used (30-33). This method is based on separation according to the hydrophobic association of folic acid compounds of various glutamate chain lengths with a C18 column. A calibration curve was made for 0–40 µg folic acid/mL (R² = 0.9999). Unexpectedly, the Florida HPLC method found an even higher [¹³C₁₁]:[¹³C₆]folic acid, ie, monoglutamyl:polyglutamyl folic acid (1.96).

HPLC or microbiological analyses cannot disclose any unlabeled monoglutamyl or hexaglutamyl folic acid in the capsules. Therefore, after deconjugation of all folic acid in the capsules to the monoglutamyl form, [¹³C₁₁]:[¹³C₆]folic acid in capsules was measured by using the Wageningen LC-MS/MS method; this method found a ratio of 1.18. Capsule analysis was not performed with the Chicago LC-MS/MS method.

Data collection
Subjects consumed one capsule with breakfast and one with dinner each day for 4 wk. Subjects kept a diary to monitor compliance with the study instructions. Blood was collected in EDTA-containing vials by venipuncture on the morning after an overnight fast on days 0, 1, 2, 4, 8, 14, and 28 during capsule intake and at 1 and 4 wk thereafter.

Dietary folate intake was assessed by a 24-h recall method. A trained dietitian contacted all participants by telephone on 5 different days of the week during the last 2 wk of the study and asked them about their food intake on the previous day. Folate intake was calculated by using the current Dutch Food Composition Table (34).

Biochemical measurements
Samples of whole blood were centrifuged for 10 min at 2600 x g and 4 °C within 30 min after the blood was drawn. Plasma was separated in 2-mL vials and stored immediately at –80 °C until further analysis. Plasma samples were transported on dry ice to laboratories and were still frozen on arrival. LC-MS/MS measurements were performed as described. Serum vitamin B-12 concentrations were assessed by using a chemiluminescence immunoassay analyzer (Immulite 2000; Diagnostic Products Company, Los Angeles, CA) at the General Practitioner's Laboratory (Velp, Netherlands).

Calculations and statistics
Enrichment levels were calculated as the concentration of the labeled compound divided by the total concentration of both labeled and unlabeled folate and expressed as a percentage. The bioavailability of hexaglutamyl folic acid relative to that of monoglutamyl folic acid was calculated by using the following equation:

RESULTS  
Data from one subject were discarded because of the use of antiinflammatory medication that could interfere with folate metabolism. The other 19 subjects (5 men and 14 women; As measured by using the Wageningen LC-MS/MS method, on day 0, no [¹³C₁₁] or [¹³C₆]5-MTHF was detected in the plasma samples. Concentrations of [¹³C₁₁] and [¹³C₆]5-MTHF in plasma increased over time until mean concentrations of 690 pmol/L (95% CI: 580, 800 pmol/L) and 440 pmol/L (360, 520 pmol/L), respectively, were reached, on day 28 (Figure 1). Labeled folate was still detected at both 1 wk and 4 wk after cessation of supplementation. The degree of folate enrichment in plasma on day 28 were 6.3% (5.6, 7.0%) for [¹³C₁₁]5-MTHF and 4.0% (3.5, 4.5%) for [¹³C₆]5-MTHF. As can be seen from Figure 2, a constant [¹³C₆]:[¹³C₁₁]5-MTHF concentration was established by day 4. On day 28, the mean ± SE ratio was 0.66 ± 0.04 (95% CI: 0.58, 0.74). Absolute concentrations of labeled and unlabeled 5-MTHF in plasma obtained by using the Wageningen LC-MS/MS method were significantly (P < 0.01) lower than those obtained by using the Chicago LC-MS/MS method. However, [¹³C₆]:[¹³C₁₁]5-MTHF obtained by either laboratory did not differ significantly from day 2 to day 28: P values ranged from 0.28 to 0.94.

View larger version (14K):
FIGURE 1.. Mean (95% CI) concentrations of [¹³C₁₁] and [¹³C₆]5-methyltetrahydrofolate (MTHF) ( and , respectively) in plasma as measured by the Wageningen liquid chromatography tandem mass spectrometry method. n = 19.

 

View larger version (17K):
FIGURE 2.. Mean (±SE) ratios of [¹³C₆] to [¹³C₁₁]5-methyltetrahydrofolate (MTHF) in plasma as measured by the Wageningen liquid chromatography tandem mass spectrometry method.

 
Finally, we calculated the bioavailability of polyglutamyl relative to that of monoglutamyl folic acid by using [¹³C₆]:[¹³C₁₁]5-MTHF in plasma with a correction for the ratio of the administered amounts of the 2 labeled compounds as assessed by different methods. On the basis of LC-MS/MS measurement of [¹³C₁₁]:[¹³C₆] folic acid in the capsules (1.18), the bioavailability of polyglutamyl folic acid relative to monoglutamyl folic acid was estimated to be 1.18 x 0.66%, ie, 78% (68, 87%). The bioavailability was higher when calculations were based on other methods for capsule content measurement (Table 1).

View this table:
TABLE 1. Analysis of monoglutamyl and hexaglutamyl folic acid in the capules and the subsequent corrections of bioavailability of hexaglutamyl relative to monoglutamyl folic acid¹

 

DISCUSSION  
We aimed to develop a dual-isotope-labeling method for studying the bioavailability of hexaglutamyl folic acid relative to monoglutamyl folic acid in humans by using multiple, orally administered low doses. We were able to measure [¹³C₆]:[¹³C₁₁]5-MTHF in plasma within narrow confidence limits, and the results were similar between the 2 laboratories using different LC-MS/MS methods. Estimation of the relative bioavailability was complicated by the fact that the doses measured in the capsules differed from the intended doses and that they varied according to the different analytic methods used. Hence, the bioavailability of polyglutamyl folic acid relative to that of monoglutamyl folic acid could only be estimated to be 78%.

In all of our human intervention studies, we verified the amount of folic acid in the fabricated capsules to ascertain whether they match the intended amounts. In supplementation studies, we usually adjust the observed changes in plasma concentration on the basis of measured doses. Unfortunately, we had underestimated how difficult it would be to analyze the amounts of monoglutamyl and hexaglutamyl folic acid in the capsules. We aimed for an amount of 50 nmol of each of the labeled folic acid compounds in the capsules. However, the measured amounts of monoglutamyl and, especially, hexaglutamyl folic acid in the capsules were lower than intended, and they differed between methods.

The discrepancies between methods may have several causes. First, the folic acid in the capsules may have been partially unlabeled, which would lead to different results from measuring labeled (LC-MS/MS) or total (HPLC, microbiological) amounts of folic acid. However, no significant signals were seen for unlabeled or partially labeled monoglutamyl or hexaglutamyl folic acid in ultraviolet LC-MS/MS spectra in the raw material. Second, deconjugation of hexaglutamyl folic acid before the analysis may have been incomplete, despite 100% deconjugation of a triglutamyl folic acid control. If this were the case, the Florida HPLC method would measure higher amounts of hexaglutamyl folic acid and therefore estimate a lower monoglutamyl:polyglutamyl folic acid in the capsules than would the Wageningen HPLC method. However, the ratio derived by using the Florida HPLC method was higher (1.96) than that derived by using the Wageningen HPLC method (1.37). Third, solubility of the folic acid compounds could have been incomplete because the capsules were dissolved in Ches/HEPES buffer at a fairly neutral pH. Therefore, we experimented with dissolving capsules in 0.05 mol borate buffer/L at a pH of 9.3, and the folic acid was dissolved during a 10-min period in an ultrasonic bath. This procedure, however, did not increase the amount of folic acid encountered in the capsules and hence did not improve solubility. In our view, these analytic inconsistencies form part of a larger problem with respect to folate analyses in foods and supplements. Discrepancy in food folate analyses between laboratories and methods has been reported previously (35-39) and has not yet been resolved.

For the current study, we considered LC-MS/MS to be the best method for capsule analysis because it takes into account the presence of any unlabeled folic acid in the capsules. Because we measured ¹³C-labeled folate in plasma, its measurement should be related to the ¹³C-labeled folic acid present in the capsules and not to any unintended, unlabeled folic acid. Indicators for the quality of Wageningen LC-MS/MS measurements were comparable to those for measurements of both HPLC methods, whereas the quality of the microbiological data was worse (Table 1). Apart from this, both microbiological and Florida HPLC methods gave rise to relative bioavailability measures of >100%, whereas the bioavailability of hexaglutamyl folic acid is not likely to exceed that of monoglutamyl folic acid, as shown by many studies (14, 15, 17, 19-22, 24, 40, 41).

LC-MS/MS measurements of absolute concentrations of labeled 5-MTHF in blood plasma were consistently higher when measured by using the Chicago LC-MS/MS method. These higher concentrations are probably due to the use of acetonitrile in the sample preparation, which releases protein-bound folate. However, the ratio of the 2 labeled compounds in plasma measured by both methods was not significantly different. Therefore, we conclude that [¹³C₆]:[¹³C₁₁]5-MTHF in plasma can be measured accurately by either of these 2 LC-MS/MS methods.

Several groups of researchers previously reported the use of a dual-label stable-isotope method for studying folate kinetics and bioavailability in humans. Investigators at the University of Florida (Gainesville, FL) evaluated a dual-label stable-isotope method in which the enrichment in plasma folate in response to an oral dose of 1010 nmol [¹³C₅]folic acid was compared with that of an intravenous dose of 226 nmol [²H₂]folic acid (42). Despite the high oral dose, plasma isotopic enrichment was 15–20 times that of injected folic acid. Therefore, it was concluded that plasma kinetics would be of limited usefulness in assessing the relative bioavailability of nutritionally relevant oral doses of labeled folic acid in single-dose experiments. Another research group also used an oral and intravenous dual-label stable-isotope method to measure the bioavailability of folic acid from fortified cereal grain foods but also found that labeled folic acid compounds were handled differently in the body because of the route of administration (43, 44). In the current study, both of the labeled folic acid compounds were administered orally, the administered dose of polyglutamyl folic acid was only 15% of the RDA (= 300 µg/d), and no preloading dose was given to subjects. Therefore, disturbance of normal folate metabolism was minimal, and cross-subject variation was small. Thus, plasma concentrations can be used to measure the relative bioavailability of folate compounds provided that a multiple-dosing scheme is being used.

Until fairly recently, techniques for folate analysis lacked the sensitivity to detect plasma concentrations of folate as low as the concentration required for our study (28, 45). Our previous target was to be able to measure enrichment levels of 1%. However, plasma enrichments of both labeled compounds were well above this target. Furthermore, [¹³C₆]:[¹³C₁₁]5-MTHF reached a plateau by 4 d. Therefore, a shorter intervention period and lower doses would be sufficient to obtain reliable data as far as measurement of plasma enrichment is concerned.

In conclusion, the dual-isotope-labeling method that we describe here provides a sensitive, accurate, and efficient method of measuring folate bioavailability, provided that doses of labeled compounds can be reliably assessed. At relevant dietary doses, the bioavailability of polyglutamyl folic acid relative to that of monoglutamyl folic acid is 78%. This method should also be suitable for the accurate measurement of the relative bioavailability of other folate species. Finally, the performance of microbiological, HPLC, and LC-MS/MS methods of folate analysis of foods and supplements should be assessed and compared.

ACKNOWLEDGMENTS  
We thank all the subjects who participated in the trial. We are grateful to Johan Lugtenburg (Leiden University) and Rob van der Steen and Marcel Bartels (ARC Laboratories) for their ingenuity in producing the labeled compounds. We also thank Frans Russel for his input with respect to the development of the method, Els Siebelink and Annelies Rotteveel for their performance of the 24-h recalls and calculation of dietary folate intake, and Peter Hollman and Diny Venema for their input with respect to the analytic issues.

AM-B, CEW, PV, MBK, and FJK were responsible for the design of the study and the data analysis; AM-B was responsible for collecting the data; JAvR, RBvB, JJPL, and SDG were responsible for the development of the liquid chromatography tandem mass spectrometry methods and analyses; all authors contributed to the writing of the manuscript. None of the authors had a personal or financial conflict of interest.

REFERENCES  

1. Herbert V. Hematology and the anemias. In: Schneider HA, Anderson CE, Coursin DB, eds. Nutritional support of medical practice. Hagerstown, MD: Harper & Row Medical Department,1982 :386 –409.
2. Smithells RW, Sheppard S, Schorah CJ. Vitamin deficiencies and neural tube defects. Arch Dis Child1976; 51 :944 –9.
3. The Homocysteine Studies Collaboration. Homocysteine and risk of ischemic heart disease and stroke: a meta-analysis. JAMA2002; 288 :2015 –22.
4. Choi SW, Mason JB. Folate status: effects on pathways of colorectal carinonsis. J Nutr2002; 132 :2413S –8S.
5. Clarke R, Smith AD, Jobst KA, Refsum H, Sutton L, Ueland PM. Folate, vitamin B₁₂ and serum total homocysteine levels in confirmed Alzheimer disease. Arch Neurol1998; 55 :1449 –55.
6. Brouwer IA, Van Dusseldorp M, West CE, Steegers-Theunissen RPM. Bioavailability and bioefficacy of folate and folic acid in humans. Nutr Res Rev2001; 14 :267 –93.
7. Sauberlich HE, Kretsch MJ, Skala JH, Johnson HL, Taylor PC. Folate requirement and metabolism in nonpregnant women. Am J Clin Nutr1987; 46 :1016 –28.
8. Schuster O, Weimann HJ, Muller J, Menke A, Menke G. Pharmacokinetics and relative bioavailability of iron and folic acid in healthy volunteers. Arzneimittelforschung1993; 43 :761 –6.
9. Seyoum E, Selhub J. Properties of food folates determined by stability and susceptibility to intestinal pteroylpolyglutamate hydrolase action. J Nutr1998; 128 :1956 –60.
10. Melse-Boonstra A, De Bree A, Verhoef P, Bjørke-Monsen AL, Verschuren WMM. Dietary monoglutamate and polyglutamate folate are associated with plasma folate concentrations in Dutch men and women aged 20–65 years. J Nutr2002; 132 :1307 –12.
11. Konings EJM, Roomans H, Dorant E, Goldbohm R, Saris W, van den Brandt P. Folate intake of the Dutch population based on newly established liquid chromotagrophy data for foods. Am J Clin Nutr2001; 73 :765 –76.
12. Gregory JF III. The bioavailability of folate. In: Bailey LB, ed. Folates in health and disease. New York, NY: Marcel Dekker Inc,1995 :195 –235.
13. Gregory JF III. Bioavailability of folate. Eur J Clin Nutr1997; 51 :S54 –9.
14. Tamura T, Stokstad ELR. The availability of food folate in man. Br J Haematol1973; 25 :513 –32.
15. Godwin HA, Rosenberg IH. Comparative studies of the intestinal absorption of [³H]pteroylmonoglutamate and [³H]pteroylheptaglutamate in man. Gastroenterol1975; 69 :364 –73.
16. Halsted CH, Baugh CM, Butterworth CEJ. Jejunal perfusion of simple and conjugated folates in man. Gastroenterology1975; 68 :261 –9.
17. Halsted CH, Reisenauer AM, Shane B, Tamura T. Availability of monoglutamyl and polyglutamyl folates in normal subjects and in patients with coeliac sprue. Gut1978; 19 :886 –91.
18. Gregory JF, Toth JP. Chemical synthesis of deuterated folate monoglutamate and in vivo assessment of urinary excretion of deuterated folates in man. Anal Biochem1988; 170 :94 –104.
19. Bailey LB, Barton LE, Hillier SE, Cerda JJ. Bioavailability of mono and polyglutamyl folate in human subjects. Nutr Rep Int1988; 38 :509 –18.
20. Keagy PM, Shane B, Oace SM. Folate bioavailability in humans: effects of wheat bran and beans. Am J Clin Nutr1988; 47 :80 –8.
21. Gregory JF, Bhandari SD, Bailey LB, Toth JP, Baumgartner TG, Cerda JJ. Relative bioavailability of deuterium-labeled monoglutmayl and hexaglutamyl folates in human subjects. Am J Clin Nutr1991; 53 :736 –40.
22. Wei MM, Bailey LB, Toth JP, Gregory JF III. Bioavailability for humans of deuterium-labeled monoglutamyl and polyglutamyl folates is affected by selected foods. J Nutr1996; 126 :3100 –8.
23. Boddie AM, Dedlow ER, Nackashi JA, et al. Folate absorption in women with a history of neural tube defect-affected pregnancy. Am J Clin Nutr2000; 72 :154 –8.
24. Melse-Boonstra A, West CE, Katan MB, Kok FJ, Verhoef P. Bioavailability of heptaglutamyl relative to monoglutamyl folic acid in healthy adults. Am J Clin Nutr2004; 79 :424 –9.
25. Melse-Boonstra A, Verhoef P, West CE. Quantifying folate bioavailability: a critical appraisal of methods. Curr Opin Clin Nutr Metab Care2004; 7 :539 –45.
26. van Lieshout M, West CE, Muhilal, et al. Bioefficacy of ß-carotene dissolved in oil studied in children in Indonesia. Am J Clin Nutr2001; 73 :949 –58.
27. Maunder P, Finglas PM, Mallet AI, et al. The synthesis of folic acid, multiply labelled with stable isotopes for bioavailability studies in human nutrition. J Chem Soc Perkin Trans1999; 1 :1311 –23.
28. Garbis SD, Melse-Boonstra A, West CE, van Breemen RB. Determination of folates in human plasma using hydrophilic interaction chromatography-tandem mass spectrometry. Anal Chem2001; 73 :5358 –64.
29. Konings EJM. A validated liquid chromoatographic method for determining folates in vegetables, milk powder, liver, and flour. J AOAC Int1999; 82 :119 –27.
30. Day BPF, Gregory JF III. Determination of folylpolyglutamate carboxypeptidase (folacin conjugase) activity by reverse phase high performance liquid chromatography. J Chromatogr1985; 318 :387 –92.
31. Bush BT, Frenz JH, Melander WR, et al. Retention behavior of pteroyl-oligo-g-L-glutamates in reversed-phase chromatography. J Chromatogr1979; 168 :343 –53.
32. Bhandari SD, Gregory JF III. Inhibition by selected food components of human and porcine intestinal pteroylpolyglutamate hydrolase activity. Am J Clin Nutr1990; 51 :87 –94.
33. Wei MM, Gregory JF III. Organic acids in selected foods inhibit porcine jejunal brush border pteroylpolyglutamate hydrolase: potential mechanism affecting the bioavailability of dietary polyglutmayl folate. J Agric Food Chem1998; 46 :211 –9.
34. Stichting Nederlands voedingsstoffenbestand. Dutch food composition table: NEVO-tabel. The Hague, Netherlands: Nutrition Centre,2001 .
35. Kariluoto S, Vahteristo L, Finglas PM, de Meer K, Nau H, Kehlenbach U. Population estimates of folate intake from food analyses. Am J Clin Nutr2002; 76 :689 –90.
36. Finglas P, Faure U, Southgate DA. First BCR-intercomparison on the determination of folates in food. Food Chem1993; 46 :199 –213.
37. Vahteristo L, Finglas P, Witthoft C, Wigertz K, Seale R, de Froidmont-Gortz I. Third EU MAT intercomparison study on food folate analysis using HPLC procedures. Food Chem1996; 57 :109 –11.
38. Finglas P, Wigertz K, Vahteristo L, Witthoft C, Southon S, de Froudmont-Gortz I. Standardisation of HPLC techniques for the determination of naturally-occurring folates in food. Food Chem1999; 64 :245 –55.
39. Puwastien P, Pinprapai N, Judprasong K, Tamura T. International inter-laboratory analyses of food folate. J Food Comp Anal2005; 18 :387 –97.
40. Bailey LB, Cerda JJ, Bloch BS, et al. Effect of age on poly- and monoglutamyl folacin absorption in human subjects. J Nutr1984; 114 :1770 –6.
41. Konings EJM, Troost FJ, Castenmiller JJM, Roomans H, van den Brandt P, Saris W. Intestinal absorption of different types of folate in healthy subjects with an ileostomy. Br J Nutr2002; 2002 :235 –42.
42. Rogers LM, Pfeiffer CM, Bailey LB, Gregory JF III. A dual-label stable-isotopic protocol is suitable for determination of folate bioavailability in humans: evaluation of urinary excretion and plasma folate kinetics of intravenous and oral doses of [¹³C₅] and [²H₂]folic acid. J Nutr1997; 127 :2321 –7.
43. Finglas PM, Witthoft C, Vahteristo L, et al. Use of an oral/intravenous dual-label stable-isotope protocol to determine folic acid bioavailability from fortified cereal grain foods in women. J Nutr2002; 132 :936 –9.
44. Finglas PM, Hart D, Wolfe C, et al. Validity of dual-label stable isotopic protocols and urinary excretion ratios to determine folate bioavailability from food. Food Nutr Bull2002; 23 (suppl):107 –12.
45. Stokes P, Webb K. Analysis of some folate monoglutamates by high-performance liquid chromatography-mass spectrometry. J Chromatogr1999; 864 :59 –67.