Previous studies have demonstrated that daily folic acid (FA) supplementation during preconception lowers the risk of neural tube defects [^1-4] and other adverse pregnancy outcomes such as low birth weight [⁵]. Because of this beneficial effect with no apparent adverse health outcomes in pregnancy, FA supplementation of approximately 400 micrograms per day (μg/d) from fortified foods, supplements, or both, was recommended for all women at risk of pregnancy [³,⁶]. To meet this recommendation, women take over-the-counter multivitamin supplements containing approximately 400 μg/d of FA [⁷]. Since clinical trials data showed that the full benefit of FA was conferred prior to conception and the national unintended pregnancy rate in the US is 49% [⁸], a more effective vehicle to reach women was pursued. In 1996, the Food and Drug Administration (FDA) approved population-wide fortification of milled grain at 140 μg FA/100 grams [⁹] to deliver an additional 100 μg/d of FA to the average adult diet [¹⁰]. However, total folate intake exceeding the recommended limit has been reported among women of child-bearing age [¹¹,¹²].

Folic acid added to foods during fortification is 70-85% bioavailable compared to 50% of folate occurring naturally in foods [⁶,¹³]. Thus, if FA supplements also are taken during pregnancy, both mother and fetus can be exposed to FA exceeding the Institute of Medicine's [⁶] recommended tolerable upper limit (TUL) of 1,000 μg/d for adult pregnant women. While folate is essential for supplying methyl groups for nucleotide synthesis and DNA replication, the effects of high FA doses are unknown. What is known is that since fortification began, American sero-surveillance data indicate that circulating folate concentrations in serum and erythrocytes have increased 50-100% in the general population [¹⁴,¹⁵]. While FA supplementation use before and during pregnancy has been monitored in numerous populations [^16-21], the extent of high FA use in pregnant women is unknown. Herein, we contribute to the monitoring effort by estimating the proportion of pregnant women reporting intake of FA doses exceeding the TUL, and determining whether socio-demographic and lifestyle factors are associated with high use within the Newborn Epigenetics STudy (NEST).

Characteristics presented in Table 1 suggest that study participants were similar to the general pregnant population in the three-county region of NC with respect to age, marital status and education of African Americans and Whites [²⁵]. However, study participants were more likely to smoke or be exposed to environmental tobacco smoke than those in the general three-county population.

Our key findings were that overall, 51% of women reported some FA intake before pregnancy and 66% reported the same during pregnancy. In addition, more than 10% of women reported daily FA supplementation exceeding the TUL before or during pregnancy. Notably, 5% of the women reported taking FA exceeding the TUL, both before and during pregnancy. FA intake exceeding TUL was most frequently reported by Caucasian women, and least frequently by those reporting at least one morbid condition. Borderline associations were also found with advanced maternal age and having a college or higher education. Recently, data from the National Health and Nutrition Examination Survey estimated that in women of childbearing age, ~1% of women are exposed to FA doses exceeding 1,000 μg/d from fortified food alone, and 26% report FA intake in doses exceeding the recommended 400 μg/d [²³].

Median folate intake among women of child-bearing age in the US is 450 μg/d (range 154-2,800ug/d) and approximately 75% is contributed by FA [¹⁰,²³]. Because synthetic folate, or FA, is 70-85% bioavailable compared to 50% of folate occurring naturally in food [⁶], these findings suggest a sizable subpopulation of mothers and fetuses may be exposed to what some describe as supraphysiological folate levels [²⁶]. To date, the adverse health effects of such exposure are unknown in humans. However, circulating folate levels have approximately doubled in the last decade [¹⁴,¹⁵]. Some large observational and randomized clinical trial data [²⁷,²⁸], but not all [²⁹], suggest an increased risk of twin pregnancies in women with high circulating serum folate levels during pregnancy, independent of age and fertility. High FA intake early in pregnancy also has been linked to an increase in the frequency of the methylenetetrahydrofolate reductase (MTHFR) 677T-allele in the fetus [³⁰,³¹]. Carrying this genetic variant has been associated with chronic conditions including depression, schizophrenia, bipolar disorder, asthma, and wheezing later in life [^32-35]. Conversely, peri-conceptional FA intake >400 μg/d was recently associated with improvements in cytosine-guanine (CpG) methylation at the insulin-like growth factor (IGF2) differentially methylated region (DMR) that regulates IGF2 imprinting in children [³⁶]. Loss of imprinting at this IGF2 DMR has been associated with a higher risk of overgrowth disorders in childhood [³⁷] and colon cancer in adulthood [³⁸]. Our findings would suggest that a sizable proportion of pregnant women and their fetuses are exposed to FA exceeding the TUL. Because the original TUL was based on the potential for excessive folate to mask vitamin B12 deficiency (rare in pregnant women), and there are potential epigenetic benefits of FA supplementation, it is prudent to continue to monitor dosages of FA exposure in the population, and study its possible effects towards the goal of establishing limits based on the genotoxic and non-genotoxic effects.

The estimate of 66% FA use during pregnancy in our study is consistent with previous reports among pregnant women in the US [³⁹,⁴⁰] and other developed countries [²⁹]. While the use of FA before pregnancy by 51% of women in the current study is higher than reports from previous studies of non-pregnant women in the US, which ranged from 27% in Arkansas to 44% in Rhode Island [^39-44], it is similar to the 53% recently reported in an American six-site, hospital-based study [¹⁶]. The significance of the inverse association between morbidity and FA exceeding the TUL is unclear although it may reflect increased counseling of FA users since such conditions require frequent contact with health care providers. Associations between Caucasian race/ethnicity and high FA intake are consistent with previous reports where any FA supplementation was evaluated [³⁹,⁴⁵], and may reflect increased knowledge of B vitamin pharmacokinetics where excess intake may not be viewed as harmful. More probable is that Caucasian race/ethnicity and perhaps advanced maternal age and higher education represent greater wherewithal to access supplemental FA, suggesting a need for population-wide re-education on currently recommended FA use.

While the design is hospital-based, the distribution of demographic characteristics in this population is similar to that of the three-county region from which study participants largely arose [²⁵,⁴⁶] and are consistent with those reported by previous studies evaluating FA intake in the US and elsewhere [¹⁸,⁴⁴,⁴⁷,⁴⁸]. A limitation of our findings is that we were unable to evaluate the potential effect of unplanned pregnancy in the association between potential correlates and FA supplementation. It is possible that FA intake exceeding TUL, during the prenatal period may be influenced by a desire to compensate for non-FA intake during peri-conception. The interpretation of these findings is also limited by our inability to prospectively monitor dosages of FA before and during pregnancy. Therefore, confirmation of these findings in larger studies will require active follow-up and validation of self-reports with serial measurement of maternal erythrocyte folate concentrations and dietary assessments. Meanwhile, prudence dictates that recommended FA doses are used by all women planning a pregnancy.

In summary, while the FDA goal of consuming at least 400 μg/d of FA pre-conceptionally has been achieved by 50% of women of childbearing age, more than one-tenth of pregnant women consume daily doses that exceed the TUL with unknown effects to humans. The use of FA exceeding the TUL was associated with Caucasian race/ethnicity, advanced maternal age and a college or higher education. This study is part of a larger effort to characterize the role of early exposures on genetic and epigenetic perturbations in humans.

The authors declare that they have no competing interests.

CH conceived of the study, designed the study directed the data collection, analysis and interpretation of the data, and drafted the manuscript. APM oversaw participant recruitment in the clinic, contributed to interpretation of the results and drafting the manuscript. JMS contributed to the study design, analysis, interpretation of the data and drafting the manuscript. MRF contributed to analysis and interpretation of the data. BC performed the statistical analysis. WDW contributed to the inception of the research hypothesis and aims. JK contributed to the logistics of data collection and interpretation of results. RLJ contributed to analysis and interpretation of the data. SKM helped conceive the study, contributed to the data interpretation and drafting the manuscript.

All authors read and approved the final manuscript.

The pre-publication history for this paper can be accessed here:

http://www.biomedcentral.com/1471-2458/11/46/prepub

We thank the parents and children of the Newborn Epigenetics STudy, research nurse Tammy Bishop, and research assistants Stacy Murray, Carole Grenier, Darby Kroyer, Natasha Duggan, and Suba Narasimhan for tracing enrolled participants; data manager Francine Overcash for statistical assistance, and the obstetrics faculty and staff at Duke University and Durham Regional Hospitals, Durham, NC.

This work was supported by the US National Institutes of Health R21ES014947, R01ES016772, K01CA104517 and the O'Keefe Foundation.