Case-control studies during the 1980s and 1990s suggested that women who took multivitamins containing folic acid during the periconceptional period had a lower risk of neural tube defects (NTDs) in their offspring. 1–5 In a large prospective study, we found a 64% reduced risk of a NTD in the offspring of women using multivitamins during the first 6 weeks of pregnancy compared with those who did not; the offspring of those whose multivitamin contained folic acid had a 73% reduced risk, whereas the offspring of those whose multivitamin did not contain folic acid had only a 7% reduced risk. 6 The epidemiological evidence was strengthened by a nonrandomized intervention study carried out in China. 7 Periconceptional use of daily supplements containing 400 μg of folic acid led to a 79% reduction in NTD risk in northern China (where there were higher baseline rates) and a 41% reduction in the South (with much lower baseline rates).
Some randomized clinical trials have also demonstrated a protective effect of folic acid supplements on NTDs. 8–12 Although none of these trials looked directly at the effects of different doses of folic acid, a high-dose trial carried out by the Medical Research Council (MRC) provided strong evidence for an overall reduction in NTD risk associated with folic acid use among women with a previous pregnancy affected by an NTD. 13 That trial used a supplementation regime of 4000 μg per day in all folic acid intervention arms. A factorial design allowed the assessment of folic acid alone, multivitamins alone (without folic acid) and folic acid plus multivitamins, compared with control group women who were given a supplement with iron and calcium alone. The occurrence of NTDs was similar in both folic acid arms of the trial, with an overall reduction in risk of 72% associated with the use of these supplements. The MRC and China studies have recently been criticized on methodological grounds, 14 but most observers agree that the overall evidence strongly supports a protective effect of folic acid-containing supplements during the periconceptional period. Questions remain, however, about the relative effectiveness of various folic acid doses. 15–17
A few of the previously published case-control studies examined folic acid dose and NTD risk 2,4,5, but only one of them found a clear dose-response relation. No prospective study or clinical trial has evaluated NTD risk according to the maternal dose of folic acid supplements or total folate intake. To address this question, we have used detailed dosage data from our earlier prospective study 6 to examine the effect of folic acid dose during the first 5 weeks of pregnancy.
In the mid-1980s, 24,559 women in the early second trimester of pregnancy were asked to participate in a prospective study of early pregnancy exposures and pregnancy outcome. The women were predominantly from the northeastern United States, and all had either a serum alpha-fetoprotein screening test or an amniocentesis. Recruitment details have been published previously. 6 The current analyses are based on data from the 23,228 women who agreed to participate and who had complete interview and outcome information for all variables of interest. The study was approved by the Institutional Review Board of the Boston University Medical Center, and all study subjects gave informed consent for participation.
All interviews were carried out by telephone by a nurse interviewer generally between the 15th and 20th gestational week. Each woman was asked to provide detailed information about her use of multivitamins and other vitamin supplements, including folic acid and yeast (which contains folic acid). The woman was asked for the brand of any supplement she took and the number of times it was taken per week. She was also asked when she had started taking the supplement and when and if there had been any change in its use.
Dietary folate intake was assessed by means of a 50-item food frequency questionnaire that was adapted from a previously validated standardized food frequency instrument. 18 The 50-item questionnaire was developed by the same investigators who designed and tested the original instrument. Food items on the checklist were specifically chosen to provide complete and valid estimates of dietary folate and included such items as liver, dark green leafy vegetables, legumes (eg, beans, lentils), other vegetables such as brussel sprouts, broccoli and peas, pizza, beef, orange juice, tomato sauce and juice, wheat germ and breakfast cereals. The woman was asked about her intake of each food item during the first 8 weeks of pregnancy. Nutrient composition was obtained for each food using a database provided by the Center for Clinical Computing at Harvard University.
The presence of fetal and congenital malformations was determined using data from the amniocentesis and ultrasound test results as well a questionnaire sent to the delivering physician at the time of expected delivery. The questionnaire asked for details of any other prenatal studies, maternal illnesses during the pregnancy, other pregnancy complications, occurrence of fetal or neonatal deaths, newborn complications and congenital malformations. All congenital anomalies were coded using the 6-Digit Code List for Reportable Congenital Anomalies published by the Centers for Disease Control and Prevention. 19
To account for differences in bioavailability of folate from different sources, we converted all reported folate intake to dietary folate equivalent (DFE) units. 20 Conversions were done as follows: (1) 1.0 μg folate from food = 1 DFE, (2) 0.5 μg folate from supplements = 1 DFE and (3) 0.6 μg folate from fortified grains = 1 DFE. Thus, for example, 400 μg of naturally occurring folate from food would provide 400 DFEs, a 400 μg vitamin supplement of folic acid would provide 800 DFEs and 400 μg of folic acid from supplemented breakfast cereals would provide about 668 DFEs.
We estimated folate dose from vitamin supplements by combining vitamin composition information from a computer file (which had detailed vitamin and mineral composition data from the mid-1980s for approximately 1,500 vitamin brands) with the reported information on frequency of intake. We used questionnaire data on starting and stopping weeks for each supplement to calculate supplemental folate dose during each week of the first trimester. Folate intake from foods (including supplemented foods) was estimated using data on usual portions and reported frequency of consumption. Thus, we were able to estimate average folate intake from foods during the first 2 months of pregnancy, whereas for supplemental folate we were able to derive estimates of intake for each individual week during the first trimester. After converting all intakes to standard DFE units, we combined weekly data from vitamin supplements with average folate intake from foods to estimate total daily folate intake during the first 8 weeks since the last menstrual period.
We examined the prevalence of NTDs according to the daily dose of folate from foods, supplements and the two combined (total folate). We conducted multiple logistic regression analyses to estimate the relative risk (prevalence ratio) of NTDs, adjusting for potential confounding by the following factors: history of NTD in any first degree relative of the offspring, maternal age, education, body mass index, history of diabetes mellitus, parity (number of previous pregnancies lasting at least 20 weeks), first trimester exposure to spermicides, fever, alcohol (drinks per week), smoking (cigarettes per day) and the intake of supplemental zinc (milligrams per day) during the first 5 weeks.
Finally, to depict the dose-response relation of total folate intake from each source with NTD risk, we used a multivariable restricted regression spline analysis, with second-order terms (quadratic) added to smooth the curves across the dose distribution. 21 A spline analysis fits separate curves for segments of the dose distribution, which allows the overall curve to reflect more accurately the shape of a dose-response trend.
For most analyses, we have restricted our examination of folate intake to the first 5 weeks since the last menstrual period. The reason for this restriction was to increase the likelihood that the woman was taking the reported dose before closure of the neural tube (between the 26th and 29th gestational day, which roughly translates to between the 40th and 43rd day since the last menstrual period). There is undoubtedly error in the precise reporting of weeks since last menstrual period, and some women’s report of pregnancy duration may also have been influenced by the knowledge of ultrasound gestational age. Finally, in this cohort, week 6 was the median reported week in which prenatal vitamins were started by women who had not previously been taking multivitamins. Given the potential for error in reporting pregnancy duration, we chose to use week 5 as the cut-off point for folate intake before closure of the neural tube.
Women using the highest dose supplements (≥800 DFEs per day) in the first 5 weeks of pregnancy were older, more likely to be college-educated and more likely to have been using multivitamins before pregnancy (Table 1). They were also more likely to have a history of diabetes mellitus or a family history of NTDs compared with women in other dose categories. Those using no folic acid had lower education levels, were slightly more likely to be obese and were more likely to have smoked cigarettes during the first trimester.
First we examined the effect of folate intake from supplements alone. Table 2 shows the prevalence of NTDs according to the daily dose of folic acid from vitamin supplements during weeks 1–5 of the pregnancy. There was no evidence of a dose-response relation according to intake from supplements alone. The offspring of women using supplemental folate had an adjusted reduction in NTD risk ranging from 44% to 71% (prevalence ratio [PR] from 0.56 to 0.29).
Then we examined the effect of folate intake from foods. Women with the lowest intakes of dietary folate from foods (<100 DFEs per day) had the highest risk of their offspring having an NTD (4.0 cases per 1,000) (Table 3). Although the PR estimates support the notion that higher food folate intakes are protective, there was again no evidence that this protective effect was linear.
Finally, we examined NTD risk according to total folate dose during weeks 1–5, as shown in Table 4. In the lowest intake category (0–149 DFEs per day), the median total folate dose consumed was 120 DFEs per day, and the prevalence of NTDs was 3.4 cases per 1,000. NTD risk generally declined as total folate dose increased. Those in the highest intake category (≥1,200), whose median total folate dose was 1,555 DFEs per day, had a risk of NTDs in their offspring of 0.8 cases per 1,000, reflecting a 77% reduction in risk. Compared with women in the lowest folate intake category, those consuming between 150 and 800 DFEs per day had a 30–34% reduced risk of NTDs, whereas those consuming between 800 and 1,200 DFEs per day had a 56% lower risk. The P-value for linear trend across these categories was 0.016. We used a linear regression model, with the median total folate dose in each category as the exposure, to assess the magnitude of the trend of decreasing NTD risk associated with increasing folate intake. We found that the prevalence of NTDs decreased by 0.78 cases (95% confidence interval [CI] = 0.47–1.09) per 1,000 pregnancies for each additional 500 DFEs per day.
Figure 1 displays the smoothed dose-response relation of total folate intake and NTD risk as estimated by the linear spline regression model. The two curves represent mean folate intake during two time periods: (1) average intake during weeks 1–5 and (2) average intake at the time of conception (ie, mean total dose of folate through the first 2 weeks since the last menstrual period). The risk of NTDs declined linearly in relation to average total folate dose during the first 5 weeks of pregnancy. The small number of cases among women consuming more than 1,000 DFEs makes it impossible to extrapolate the shape of the dose-response curve at higher doses. Similarly, the small numbers of women with very low total folate intakes (as shown in Table 4) provides less confidence in the shape of the dose-response curve in that region as well. The risk of NTDs associated with folate intake at the time of conception declined steeply between 600 and 1,000 DFEs per day and then seemed to level off at about 1,000 DFEs per day. Again, we had insufficient data at higher doses to draw firm conclusions about the shape of the curve above 1,000 DFEs per day at the time of conception.
Finally, we carried out a subanalysis designed to assess whether women who were taking supplemental folate before conception had any greater benefit from its use than did women who started taking supplements in the first few weeks following conception. Among the 7,044 women who were taking a vitamin containing folic acid of any dose before or during the first 2 weeks since the last menstrual period (ie, before conception), six NTD cases occurred, reflecting a 76% reduction in risk (PR = 0.24; CI = 0.09–0.66) compared with those who started supplements later. Of the 16,184 women not taking any folic-acid containing multivitamins until week three or later, 3,028 women first started supplements during weeks 3 through 5. The six additional NTD cases that occurred among these women reflected a 44% lower risk of a NTD (PR = 0.56; CI = 0.21–1.5) compared with the risk seen among women not starting folate supplementation until week 6 or later.
In this study, we combined folate intake data from vitamin supplements and foods. We converted folate intake from all sources to DFE units to account for the differing bioavailability of folate from different sources. We found that NTD risk generally declined with increasing total daily folate intake during the first 5 weeks since the last menstrual period (Table 4 and Figure 1). This effect was even stronger for those consuming more than 600 DFEs per day of total folate at the time of conception. There was no progressive dose-response relation for either vitamin supplements alone or folate intake from foods alone with regard to NTD risk. This result for supplements alone is not surprising because the use of vitamin supplements containing folic acid might well have a different effect among women whose diet is already rich in folate, compared with women whose dietary folate intake is very low. Because the diet questionnaire targeted average intake in the first 8 weeks of pregnancy, the absence of a dose-response relation for food may have resulted from imprecise estimation of folate consumed from food sources specifically in the weeks before neural tube closure. It may also be that diets that are low in naturally occurring folate may be less harmful for women using a folic acid-containing multivitamin at the time of conception.
For the most part, folate dose information in this prospective study was gathered before the outcome of the pregnancy was known. However, 19 of the 49 women with a NTD offspring were aware of the prenatal test results before the interview. Nonetheless, the study was carried out before widespread publicity about the potential protective effects of folic acid, making biased reporting less likely. In addition, the close proximity of the interview to the exposure period would have enhanced the accuracy of the subject’s report of exposure information. The study is further strengthened by the fact that we gathered detailed information on folic acid supplement use as well as intake of folate-containing foods, allowing us to estimate total folate dose for each week during early pregnancy.
Few women were taking high-dose folic acid supplements in this study, and so we are unable to evaluate the effects of high-dose supplementation levels. In fact, only five women in this cohort took supplements of 4 mg per day (8,000 DFEs) during the first 5 weeks, as used in the MRC study. 13 The relative reduction in risk associated with total folate doses of approximately 800 DFEs per day in our study was similar in magnitude to that seen with the higher-dose supplements used in the MRC trial. However, this comparison is complicated by the fact that this latter study targeted a high-risk population (ie, those with previous NTD-affected pregnancy).
In 1996, the weight of evidence linking folate intake with NTD risk led the United States Food and Drug Administration to embark on a program designed to add 100 μg (167 DFEs) of supplemental folate per day to the diets of all Americans by adding 140 μg of folic acid to every 100 g of grain. 22 Similar programs are being considered in the United Kingdom and elsewhere. In the meantime, the debate continues about the optimal levels of fortification, with some arguing that the current level of fortification in the U.S. is too low 23,24 and others calling for restraint until the efficacy of the current program is tested and more data on safety of high-dose folate in the general population are available. 16,17,25 Recent data from the Centers for Disease Control suggest that since the onset of mandatory food fortification in the U.S., there has been a 13% decline in the occurrence of NTDs among the offspring of women who received only late (third trimester) or no prenatal care. 26 There is evidence that the current levels of food fortification have dramatically raised blood levels of folate in the population at large, 27,28 leading to additional concerns about the safety of still higher levels of fortification. And at the same time, there is some disappointment that the NTD risk reduction has not been greater.
In summary, there are still not enough data to determine the optimal periconceptional folate intake needed to prevent NTDs. However, this study provides important additional evidence that NTD risk declines steeply with increasing intake of modest amounts of total folate in the early weeks of pregnancy. These results highlight the need to consider total folate dose rather than supplemental folate intake alone in establishing public health guidelines for NTD prevention.
We are indebted to the staff of Boston Collaborative Drug Surveillance Program, who oversaw the initial data collection, and to Cristina Cann for her thoughtful assistance in the coding of pregnancy outcomes.
1. Mulinare J, Cordero JF, Erickson JD, Berry RJ. Periconceptional use of multivitamins and the occurrence of neural tube defect. JAMA 1988; 260: 3141–3145.
2. Bower C, Stanley FJ. Dietary folate
as a risk factor for neural tube defects
: evidence from a case-control study in Western Australia. Med J Aust 1989; 150: 613–619.
3. Winship KA, Cahal DA, Weber JCP, Griffin JP. Maternal drug histories and central nervous system anomalies. Arch Dis Child 1984; 59: 1052–1060.
4. Werler MM, Shapiro S, Mitchell AA. Periconceptional folic acid
exposure and risk of occurrent neural tube defects
. JAMA 1993; 269: 1257–1261.
5. Shaw GM, Schaffer D, Velie EM, Morland K, Harris JA. Periconceptional vitamin use, dietary folate
, and the occurrence of neural tube defects
. Epidemiology 1995; 6: 219–226.
6. Milunsky A, Jick H, Jick SS, et al
. Multivitamin/folic acid
supplementation in early pregnancy reduces the prevalence of neural tube defects
. JAMA 1989; 262: 2847–2852.
7. Berry RJ, Zhu L, Erickson JD, et al
. Prevention of neural tube defects
with folic acid
in China. The China-U.S. Collaborative Project for Neural Tube Defect Prevention. N Engl J Med 1999; 341: 1485–1490.
8. Laurence KM, James N, Miller MH, Tennant GB, Campbell H. Double-blind randomised controlled trial of folate
treatment before conception to prevent recurrence of neural-tube defects. BMJ 1981; 282: 1509–1511.
9. Smithells RW, Sheppard S, Wild J, Schorah CJ. Prevention of neural tube defect recurrences in Yorkshire: final report. Lancet 1989; 2: 498–499.
10. Vergel RG, Sanchez LR, Heredero BL, Rodriguez PL, Martinez AJ. Primary prevention of neural tube defects
with folic acid
supplementation: Cuban experience. Prenat Diagnos 1990; 10: 149–152.
11. Nevin NC, Seller MJ. Prevention of neural-tube-defect recurrences. Lancet 1990; 335: 178–179.
12. Czeizel EA, Dudas I. Prevention of the first occurrence of neural-tube defects by periconceptional vitamin supplementation. N Engl J Med 1992; 327: 1832–1835.
13. MRC Vitamin Study Research Group. Prevention of neural tube defects
: results of the Medical Research Council Vitamin Study. Lancet 1991; 338: 131–137.
14. Turner LA, Morrison H, Prabhakaran VM. Do we need another randomized controlled trial of folic acid
alone? Epidemiology 2001; 12: 262–265.
15. Botto LD, Moore CA, Khoury MJ, Erickson JD. Medical progress: neural-tube defects. N Engl J Med 1999; 341: 1509–1519.
16. Moore LL. Is the jury still out on folic acid
and congenital anomalies? Epidemiology 2001; 12: 141–144.
17. Mills JL. Fortification of foods with folic acid
– how much is enough? N Engl J Med 2000; 342: 1442–1445.
18. Willett WC, Sampson L, Stampfer MJ, et al
. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol 1985; 122: 51–65.
19. Centers for Disease Control and Prevention. CDC 6-digit code list for reportable congenital anomalies. In: Metropolitan Atlanta Congenital Defects Program. Surveillance Procedure Manual and Guide to Computerized Anomaly Record. Atlanta: Centers for Disease Control and Prevention, 1998.
20. The Standing Committee on the Scientific Evaluation of Dietary Reference Intakes and its Panel on Folate
, Other B Vitamins, and Choline and Subcommittee on Upper Reference Levels of Nutrients, Food, and Nutrition Board, Institute of Medicine. Dietary Reference Intakes for Thiamin, Riboflavin, Niacin, Vitamin B6, Folate, Vitamin B12, Pantothenic Acid, Biotin, and Choline
. Washington DC: National Academy Press, 2000:196–305.
21. Greenland S. Dose-response and trend analysis in epidemiology: alternatives to categorical analyses. Epidemiology 1995; 6: 356–365.
22. DHHS, FDA. Food standards: amendment of standards of identity for enriched grain products to require addition of folic acid
. Federal Register 1996; 61: 8781–8807.
23. Meyer RE, Oakley GP Jr. Folic acid
fortification [correspondence]. Lancet 1999; 354: 2168.
24. American Academy of Pediatrics. Folic acid
for the prevention of neural tube defects
. Pediatrics 1999; 104: 325–327.
25. Mills JL, England L. Food fortification to prevent neural tube defects
: is it working? JAMA 2001; 285: 3022–3023.
26. Honein MA, Paulozzi LJ, Mathews TJ, Erickson JD, Wong LY. Impact of folic acid
fortification of the U.S. food supply on the occurrence of neural tube defects
. JAMA 2001; 285: 2981–2986.
27. Jacques PF, Selhub J, Bostom AG, Wilson PWF, Rosenberg IH. The effect of folic acid
fortification on plasma folate
and total homocysteine concentrations. N Engl J Med 1999; 340: 1449–1454.
28. Lawrence JM, Chie V, Petitti DB. Fortification of foods with folic acid
[correspondence]. N Engl J Med 2000; 343: 970.