点击显示 收起
1 Department of Applied Chemistry and Microbiology University of Helsinki PO Box 27 FIN-00014 University of Helsinki Finland E-mail: susanna.kariluoto{at}helsinki.fi
2 Institute of Food Research Norwich United Kingdom
3 Department of Clinical Chemistry Vrije Universiteit Medical Center Amsterdam Netherlands
4 Department of Food Toxicology School of Veterinary Medicine Hannover Germany
Dear Sir:
In their article on the folate intakes of the adult Dutch population, Konings et al (1) presented calculations of folate intake based on the folate contents of foods determined by liquid chromatography (LC). The dietary folate intake of the Dutch population was estimated to be 182 ± 119 µg/d. The limitations of their study, and hence possible biases in the final estimates of folate intake, require comment.
In comparison with other published data on folate intakes in European populations, the intakes presented by Konings et al (1) seem low. In the study by de Bree et al (2), average folate intakes in European populations were reported to be 291 and 247 µg/d for males and females, respectively. As part of the European Union’s fifth framework project, "Folate: From Food to Functionality and Optimal Health" (3), folate intakes for populations in 7 European countries (Finland, Germany, Italy, the Netherlands, Spain, Sweden, and the United Kingdom) were estimated on the basis of each country’s food consumption and food-composition data. Dietary folate intakes ranged from 217 to 310 µg/d, the average being 285 ± 30 µg/d for males and 230 ± 20 µg/d for females.
In addition, an intercomparison study of suitable methods of analysis for food folates was organized by using 4 certified reference materials (CRMs): CRM 121 (whole-meal flour), CRM 421 (milk powder), CRM 485 (mixed vegetables), and CRM 487 (pig’s liver). Nine project participants and the National Food Administration (Sweden) analyzed the samples. To increase the number of comparable results, the data from Konings previous publication (4) were included. The number of accepted results was 7–9 for each CRM, 3–4 of which were obtained by a microbiological assay (MA) and 4–5 by HPLC. One of the HPLC laboratories used microbiological detection (LC-MA). In addition, 2 HPLC laboratories determined 5-methyltetrahydrofolate only. The certified values for total folate in these CRMs were obtained by MA by a group of experienced laboratories (5).
Our findings indicate the current status of food folate analysis in Europe. The results obtained by HPLC were 27–40% lower than the microbiological results, except for pig’s liver. Indeed, Konings et al also reported a difference of the same magnitude (23–27%) between the results by LC and those by MA. The within-laboratory variation was generally small (3–11%), whereas the between-laboratory variation was large (7–51%). Generally, folic acid and 5-methyltetrahydrofolate could be determined with satisfactory accuracy but no agreement was found for the other vitamers.
The distribution of the folate vitamers and the total folate content of CRM 485 (mixed vegetables) are presented in Figure 1. The sum of folate vitamers was 73% of the microbiologically determined folate content for mixed vegetables but only 60% for whole-meal flour. The distribution patterns from CRMs containing several vitamers (whole-meal flour and pig’s liver) varied considerably from one laboratory to another.
FIGURE 1. . The total folate content (), determined by microbiological assay (MA) in 4 laboratories and individual folate contents (, folic acid; , 5-HCO-H4-folate; , 5-CH3-H4-folate; , H4-folate) determined by HPLC (LC) in 6 laboratories of certified reference material 485 (mixed vegetables) in the European intercomparison study. The SD of the certified value is indicated by the horizontal dashed lines. The results of Konings (4) are included in LC 3. LC-MA, HPLC with microbiological detection; DM, dry matter.
There are several possible explanations for the observed discrepancy. Some of the peaks in HPLC chromatograms were often masked, which made identification and accurate quantification difficult. The existence of unidentified compounds, either folate or nonfolate, was common. It is known, for example, that 5-formyltetrahydrofolate and 10-formyltetrahydrofolate can convert to 5,10-methylenetetrahydrofolate under acidic conditions (pH < 2) used in the current HPLC mobile phases. 5,10-Methylenetetrahydrofolate does not have sufficient fluorescence to be quantified accurately. However, alternate detection systems, such as electrochemical or mass spectrometry, have been developed (6).
There is also the possibility that the bacteria used in the MA gave unequal responses to the different vitamers or can be affected by some nonfolate compounds in the samples. However, the latter was evaluated on the basis of the combined LC-MA results. The data clearly show the absence of any nonfolate peaks with organism activity.
We conclude that the MA is preferred for the determination of the total folate content of foods, especially if data on individual folates is not considered necessary. Furthermore, we stress the need for careful validation of HPLC methods. Given the considerable, yet unidentified, difference between the microbiological and HPLC data, we advise caution when estimating dietary folate intakes at the population level.
REFERENCES