|
|
||||||||
School of Public Health, University of California, Berkeley, California
Address reprint requests to: Marion Dietrich, Ph.D., HNRCA @ Tufts University, Nutritional Epidemiology, 711 Washington Street, Boston, MA 02111. E-mail: dietrich{at}tufts.edu
| ABSTRACT |
|---|
|
|
|---|
Methods: We compared data from two National Health and Nutrition Examination Surveys (NHANES): NHANES III, conducted during 1988 to 1994, reflecting the time prior to folate fortification, and NHANES 19992000, reflecting the time period after fortification.
Results: Mandatory folic acid fortification led to significant increases in both serum and erythrocyte folate concentrations in all sex and age groups. In the overall study population the mean serum folate concentration increased more than two-fold (136%), from 11.4 nmol/L to 26.9 nmol/L, and the mean erythrocyte folate concentration increased by 57 percent, from 375 nmol/L to 590 nmol/L. Less than 10% of women of childbearing age reached the recommended erythrocyte folate concentration of greater than 906 nmol/L that has been shown to be associated with a significant reduction in neural tube defect (NTD) risk. After fortification, the category "bread, rolls, and crackers" became the single largest contributor of total folate to the American diet, contributing 15.6% of total intake, surpassing vegetables, which were the number one folate food source prior to fortification. Dietary intake of total folate increased significantly in almost all sex and age groups, except in females over 60 years of age. The mean dietary total folate intake of the study population increased by 76 µg/d (28%), from 275 µg/d to 351 µg/d.
Conclusions: The fortification of enriched cereal-grain products with folic acid lead to a significant improvement of blood folate status of the overall adult, non-supplement using, US population. However, women of childbearing age may take folic acid supplements to reach erythrocyte folate levels that have been associated with decreased risk of NTDs.
Key words: folate, folic acid, total folate, food sources, blood folate, NHANES III, NHANES 19992000
| INTRODUCTION |
|---|
|
|
|---|
The aim of this study was to explore the changes in serum and erythrocyte folate status, food sources for folate, and dietary intake of total folate (including both food folate and folic acid), following folic acid fortification of enriched cereal-grain products. The purpose of this investigation was to determine whether fortification had resulted in sufficient improvement in the folate status of the U.S. adult population and women of childbearing age in particular. This was accomplished by comparing data from two National Health and Nutrition Examination Surveys (NHANES): NHANES III, conducted during 1988 to 1994, reflecting the time prior to folate fortification, and NHANES 19992000, reflecting the time period after fortification.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Both NHANES III and NHANES 19992000 randomly selected subjects using a stratified multistage probability design involving counties, blocks, and households. NHANES III was conducted in two three-year phases (19881991 and 19911994). These two separate national probability samples included a combined total of approximately forty thousand subjects aged two months or older. Young children, older people, non-Hispanic blacks, and Hispanics were over-sampled in order to obtain precise estimates of health characteristics for these groups. Each nationally representative annual sample of NHANES 1999 and NHANES 2000 included a total of approximately nine thousand subjects of all ages. NHANES 19992000 over-sampled adolescents, older people, pregnant women, non-Hispanic blacks, Hispanics, and low-income individuals. Earlier publications describe further details of the survey design and sampling methods for NHANES III [13,14] and NHANES 19992000 [14,15]. Overall, for the most part, the statistical principles and reliability considerations stated in the NHANES III Analytic Guidelines are the same for NHANES 19992000 data sets [16]. However, standard errors were estimated using SUDAAN by means of the "delete 1 jackknife (JK1) method" in contrast to the Taylor Series Linearization method that was used to estimate standard errors in previous NHANES [17,18].
Physical examinations and dietary interviews for each subject were conducted in mobile examination centers. Food intake information was obtained via a 24-hour dietary recall. Details of the procedure utilized for the dietary interviews have been published previously for NHANES III [19] and NHANES 19992000 [19,20]. During the physical examination, blood samples were obtained by venipuncture. Serum and erythrocyte folate were assessed for subjects four years and older during NHANES III and for subjects three years and older during NHANES 19992000. NHANES III utilized two methods to assess serum and erythrocyte folate concentrations. Prior to November 1993, serum and erythrocyte folate concentrations in NHANES III were measured using the Quanta Phase I Folate Radioassay Kit (Bio-Rad Laboratories, Hercules, California). After December 1993 in NHANES III, the assays were performed using the Quanta Phase II Folate Radioassay Kit (Bio-Rad Laboratories, Hercules, California). NHANES 19992000 utilized the latter method. The CDC has applied a correction factor to the earlier serum and erythrocyte folate concentrations in the NHANES III dataset. This correction accounts for the variation between the two assays that results from incorrect calibrator solutions in the Quanta Phase I Kit. The same analyst at the NHANES Central Laboratory of CDC conducted all analyses. There was no evidence of analytical drift, and the mean coefficient of variation up to 1999 was 5% [21,22]. Detailed laboratory procedures are available elsewhere for NHANES III [23,24] and NHANES 19992000 [25].
The protocols for NHANES were reviewed and approved by the Institutional Review Board of the National Center for Health Statistics.
Total Folate
The dietary folate data in the NHANES nutrient database are reported as total folate intake in µg/day. The database does not contain folate intake data in form of Dietary Folate Equivalents (DFE), which would take the higher bioavailability of folic acid compared to food folate into account (DFE; 1 DFE = 1 µg food folate = 0.6 µg folic acid from fortified food) [2,26]. Based on this, the dietary data presented here in this study are in micrograms of total folate and the term total folate refers to the combination of food folate and folic acid from fortification.
Study Subjects
Subjects for the study were participants in either NHANES III or NHANES 19992000. Subjects who had unreliable dietary recall records, as identified by NHANES, were excluded from our analysis. Pregnant women were excluded on the basis that hemodilution causes a normal decrease in serum and erythrocyte folate values [2], and subjects who reported taking any supplements within the 30 days prior to the NHANES III or NHANES 19992000 interview were excluded as well. We excluded supplement users because the supplement data in the NHANES 19992000 data release did not include specific information on the type and content as well as the amount of micronutrient supplements in general at the time this analysis was conducted. Therefore, no accurate estimation of folic acid intake from supplements would have been possible.
The NHANES III dataset contained a total of 17,158 subjects aged 18 years and older. Of those, 73 had unreliable diet records, 275 were pregnant, 1,057 were less than 20 years old, and 5,834 were supplement users. Therefore, a total of 7,239 were eliminated, generating a final sample size of 9,919 subjects. The NHANES 19992000 dataset contained a total of 8,843 subjects of all ages. Of those, 213 had unreliable diet records, 215 were pregnant, 4,363 were less than 20 years old, and 1,931 were supplement users. Thus, the exclusion of a total of 6,722 subjects lead to a final sample size of 2,121. For both the NHANES III and NHANES 19992000 databases, the final sample sizes of 9,919 and 2,121, respectively, represent those subjects for whom dietary data were available. Within the NHANES III dataset, serum folate data were available for only 9,430 of the 9,919 subjects, and erythrocyte folate data were available for 9,438 subjects. Within the NHANES 19992000 dataset, serum and erythrocyte folate data were available for 1,978 and 2,007 subjects, respectively.
Subjects were divided into three age categories (2039 years, 4059 years, and 60+ years) following the analytic guidelines for NHANES 19992000 [17]. Individuals less than 20 years of age were excluded due to conflict between the age grouping recommended by the analytic guidelines for NHANES 19992000 [17] and those used by the Food and Nutrition Board when establishing Dietary Reference Intakes for folate [2].
Statistical Analysis
All analyses were carried out using SUDAAN for IBM PC (version 8.0.2, Research Triangle Institute, Research Triangle Park, North Carolina), a data analysis package for multistage sample designs. This method accounts for different sample weights and the effects of the complex sample design on variance estimation. Thus, the analyses included an adjustment for over-sampling and non-response. Standard errors of means were calculated using the Taylor Series Linearization method for NHANES III and the Jackknife (JK1) procedure for NHANES 19992000. Statistical significance was determined using a one-tailed Z test, based on the assumption that fortification could only improve folate status. Due to the large sample size, differences were considered significant only if p
0.01.
| RESULTS |
|---|
|
|
|---|
|
|
|
|
|
|
| DISCUSSION |
|---|
With regard to a subgroup of the U.S. population, a report by the Centers for Disease Control and Prevention (CDC) was published in 2002 with results on the serum folate status in women of childbearing age, also using the NHANES III and NHANES 19992000 datasets [22]. In this CDC report, the median serum folate concentrations for women aged 1544 years increased from 4.8 to 13.0 ng/mL from NHANES III to NHANES 19992000, corresponding to 10.9 nmol/L to 29.5 nmol/L, respectively. This increase is comparable to our findings of the two age groups that cover approximately the same age range described in the CDC report. In our analysis, in females of 2039 and 4059 years of age, the medians rose from 8.8 and 9.5 nmol/L to 22.3 and 25.0 nmol/L serum folate, respectively. It is not clear from the CDC report, whether subjects reporting any kind of supplement use in the last 30 days were excluded from their analysis. Therefore, the slightly higher levels after fortification reported by the CDC could result from higher serum levels in supplement users and from differences in sample sizes and age ranges.
In the Framingham Offspring Study, mean plasma folate concentrations of males and females aged 32 to 80 increased from 11 to 23 nmol/L (p < 0.001) among subjects who did not use vitamin supplements [7]. This result is very similar to the result from the analysis presented here, where the mean concentration of all subjects increased from 11 to 27 nmol/L.
Lawrence et al. [27] analyzed serum folate data from more than 98,000 male and female subjects (age range less than 4 years to 70 years or older) of the Kaiser Permanentes Southern California database. These blood samples were collected from 1994 on through 1998, the year of full implementation of the mandatory food fortification. Lawrence et al. found that the median serum folate levels increased gradually from 28.5 nmol/L pre-fortification to 42.4 nmol/L the first year of fortification (from 12.6 µg/L to 18.7 µg/L). No information is given on subjects folic acid supplement use. Caudill et al. [28] investigated 135 Southern Californian women 1845 years of age, non-supplement users, after mandatory folic acid fortification. The authors report a mean serum folate level of 50 nmol/L. The blood folate levels reported in these two California cohorts are substantially higher compared to the other reports on post-fortification serum folate levels. Possible explanations for this difference might lie in geographic location of the study subjects and methodological issues such as different laboratory measurement methods for blood folate, and possibly the inclusion of folic acid supplement users in the Lawrence et al. study.
In summary, these post-fortification results indicate that increasing proportions of the population are now well exceeding the minimum recommended serum folate level of 7 nmol/L. With regard to the prevention of NTDs, women of childbearing age are now well exceeding the lower limit of the acceptable range of serum folate levels (
13.6 nmol/L) associated with very low risk for NTDs [29].
Red blood cell (RBC) folate is a better measure of long-term folate status than serum concentrations because this marker reflects tissue folate stores. A RBC folate level of >906 nmol/L has been associated with significantly decreased risk for pregnancies affected with neural tube defects. Metabolically, increased folate intake first increases serum folate concentrations then increases erythrocyte concentrations. Folate is incorporated into erythrocytes during their formation in the bone marrow [2]. A value of 305 nmol/L of erythrocyte folate has been chosen as the cutoff point for adequate folate status [2] for the general adult population. We found that fortification has significantly improved the median erythrocyte folate status in the total US population by increasing it by over fifty percent to 541 nmol/L (the mean levels increased to 590 nmol/L ± 11.6 (±SEM)), leaving only less than four percent of the total US population with inadequate folate RBC status. Within the two age categories of females that include women of childbearing ages (2039 and 4059 yrs), the median RBC folate increased to 505 nmol/L and 587 nmol/L, respectively (the mean levels were 556 ± 16.1 (±SEM) and 629 ± 21.6 (±SEM), respectively). More than 90% of these women did not reach RBC levels of the recommended 906 nmol/L (400 ng/mL), a level that has been shown to be associated with a significant reduction in NTD risk [30].
To date, two other studies have investigated the effect of folic acid fortification of the food supply on RBC folate status. In 2002, the U.S. Centers for Disease Control and Prevention (CDC) reported that the median RBC folate level in U.S. women of childbearing age (1544 years) increased from 363 nmol/L to 598 nmol/L (reported as 159.9 ng/mL and 263.6 ng/mL, multiplied by the conversion factor 2.2655) between NHANES III and NHANES 19992000 [22]. This result is consistent with our report, indicating that the majority of U.S. women of childbearing age do not reach the recommended RBC folate level by dietary folate intake alone.
The other recent report examining changes in RBC folate status is the report by Caudill et al. [28]. Caudill et al. investigated 135 Southern Californian women 1845 years of age, also non-supplement users, after mandatory folic acid fortification. Red blood cell mean levels post-fortification were 1307 ± 349 nmol/L (±SD), indicating that this relatively small cohort of Californian women reach the recommended RBC folate level. This RBC folate level is more than twice as high as the mean level the CDC and we found in women of comparable age in the NHANES 19992000 dataset. This larger increase in RBC folate concentration cannot readily be explained. Similar to the differences seen between studies reporting changes in serum folate levels, possible explanations for this discrepancy might lie in geographical and methodological issues, such as different dietary habits of study subjects, differences in laboratory methods used, or sample size differences. The Caudill et al. study used a much smaller sample size than those of the NHANES analyses and was of cross-sectional design, investigating post-fortification levels only. Geographically, dietary intake of vegetables and orange juice, which are both major dietary sources of food folate, might be higher in California, due to more intensified diet and health campaigns than in other parts of the country.
In summary, although the serum folate levels of women of childbearing age are well exceeding the acceptable level, that subject group does not reach the RBC folate levels associated with a significant reduction in NTD risk. RBC folate status reflects the long-term folate status and is therefore considered to be a better proxy for blood folate status. These results thus indicate that women of childbearing age should take folic acid supplements in order to reach the recommended RBC folate level. With regard to the general U.S. adult population, only a very small fraction does not reach adequate RBC folate levels.
Our analysis also included an investigation of changes in total dietary folate intake due to fortification. However, independent determination of folic acid intake from food fortification could not be performed, because the food composition tables used for the NHANES database do not distinguish between naturally occurring food folate and synthetic folic acid. Intake in dietary folate equivalents, therefore, cannot be determined. Due to this reason, we were not able to compare dietary post-fortification intakes to estimated requirement levels (EARs), which are reported in dietary folate equivalents (DFE). The total folate intakes reported in this study also consequently underestimate the post-fortification increase in folate, due to this incongruity in units. However, the observed absolute changes in folate intake are most likely mainly due to added folic acid rather than changes in food patterns.
The mean absolute increase in total folate intake in all subjects was 76 µg per day (28% increase), which is within the FDAs estimated range that fortification would supply individuals with an additional 70 to 130 µg of folic acid per day [31]. However, the changes in blood levels in our analysis are substantially higher, with 57 and 136 percent increases in serum and RBC folate concentrations, respectively. This large discrepancy is likely due to an underestimation of dietary folate intake, because it has been reported that many enriched food products may contain higher levels of folic acid than required by the Federal regulations [32]. These results highlight the existing limitations in measuring dietary folate intake in the US. Currently, no database analyzing the folic acid levels in foods is available, and information on food labels provided by manufacturers has been reported to be unreliable. Many products contain higher levels of folic acid than required by the FDA regulation [32,33].
Using our analysis of increases in blood levels, we can calculate an estimate of the change in the dietary intake after fortification, giving insight into the possible extent of the underestimation affecting our dietary data. We observed an absolute change in serum folate levels of 16 nmol/L (7 µg/L) in the overall U.S. population. Quinlivan and Gregory [34] published results from a linear regression analysis, using data from published studies, to evaluate the relationship between chronic folic acid intake and resulting changes in serum folate concentrations. These authors report that an additional 70 µg per day would lead to a change of approximately 4.3 nmol/L in serum plasma concentration. Given the correlation coefficient of r = 0.984 of their linear correlation, it can be estimated that an increase in serum folate of 16 nmol/L is associated with an additional 260 µg of folic acid from food fortification. This is approximately double the increase in folic acid intake than what the FDA anticipated (between 70 and 130 µg/d) [31]. This estimation of an additional 260 µg intake of folic acid is consistent with data in the literature. Jacques et al. [7] and Lawrence et al. [27] also suggest that dietary folate intake levels actually have increased by approximately more than 200 µg/d.
If fortification has in fact supplied the U.S. population with an additional 260 µg of folic acid per day, it raises the concern that a greater number of individuals are now exceeding the Tolerable Upper Intake Level (UL) for folic acid. Vitamin B12 deficiency may be masked by excess folic acid intake and could lead to irreparable neurological disorders. Thus, efforts need to be undertaken to evaluate the long-term effects and safety of the food fortification program, especially with regard to supplement users and the elderly. However, until a database that reports the actual food folate and folic acid content of foods becomes available, it will not be possible to accurately determine the proportion of the population exceeding the UL.
Our analysis on the changes in the ranking of food sources contributing to dietary folate intake showed, as predicted, that those foods that were fortified, such as bread, rice, and pasta, became more important sources of folic acid post-fortification.
| CONCLUSION |
|---|
|
|
|---|
While examining the changes in folate status due to fortification, this study relied primarily on blood indicators of nutrient status, due to limitations in the dietary folate intake data generated by this study. These limitations highlight an urgent need for the generation of reliable food composition databases for enriched cereal-grain products and products containing enriched cereal-grain products. These databases should be developed using data on the actual concentrations of food folate and folic acid in the foods and report the total folate concentrations in Dietary Folate Equivalents (DFE).
| ACKNOWLEDGMENTS |
|---|
|
|
|---|
| FOOTNOTES |
|---|
|
|
|---|
Received May 23, 2004. Accepted December 2, 2004.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. M. Winkels, I. A. Brouwer, P. Verhoef, F. V. A. van Oort, J. Durga, and M. B. Katan Gender and Body Size Affect the Response of Erythrocyte Folate to Folic Acid Treatment J. Nutr., August 1, 2008; 138(8): 1456 - 1461. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. B. Dowd and A. E Aiello Did national folic acid fortification reduce socioeconomic and racial disparities in folate status in the US? Int. J. Epidemiol., May 2, 2008; (2008) dyn066v1. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Morin, K. Demers, C. Giguere, D. S.-C. Tribble, and J. Lane Prise de supplements multivitaminiques pour la prevention primaire des anomalies congenitalies: Application d'une pratique clinique preventive Can Fam Physician, December 1, 2007; 53(12): 2142 - 2143. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. M Pfeiffer, C. L Johnson, R. B Jain, E. A Yetley, M. F. Picciano, J. I Rader, K. D Fisher, J. Mulinare, and J. D Osterloh Trends in blood folate and vitamin B-12 concentrations in the United States, 1988 2004 Am. J. Clinical Nutrition, September 1, 2007; 86(3): 718 - 727. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Chattopadhyay, R. Tamari, S. H. Min, R. Zhao, E. Tsai, and I. D. Goldman Commentary: A Case for Minimizing Folate Supplementation in Clinical Regimens with Pemetrexed Based on the Marked Sensitivity of the Drug to Folate Availability Oncologist, July 1, 2007; 12(7): 808 - 815. [Abstract] [Full Text] [PDF] |
||||
![]() |
Q.-H. Yang, H. K Carter, J. Mulinare, R. Berry, J. Friedman, and J D. Erickson Race-ethnicity differences in folic acid intake in women of childbearing age in the United States after folic acid fortification: findings from the National Health and Nutrition Examination Survey, 2001-2002 Am. J. Clinical Nutrition, May 1, 2007; 85(5): 1409 - 1416. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Shirodaria, C. Antoniades, J. Lee, C. E. Jackson, M. D. Robson, J. M. Francis, S. J. Moat, C. Ratnatunga, R. Pillai, H. Refsum, et al. Global Improvement of Vascular Function and Redox State With Low-Dose Folic Acid: Implications for Folate Therapy in Patients With Coronary Artery Disease Circulation, May 1, 2007; 115(17): 2262 - 2270. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Poo-Prieto, D. B. Haytowitz, J. M. Holden, G. Rogers, S. F. Choumenkovitch, P. F. Jacques, and J. Selhub Use of the Affinity/HPLC Method for Quantitative Estimation of Folic Acid in Enriched Cereal-Grain Products J. Nutr., December 1, 2006; 136(12): 3079 - 3083. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. L. Sherwood, L. A. Houghton, V. Tarasuk, and D. L. O'Connor One-Third of Pregnant and Lactating Women May Not Be Meeting Their Folate Requirements from Diet Alone Based on Mandated Levels of Folic Acid Fortification J. Nutr., November 1, 2006; 136(11): 2820 - 2826. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |