Folic Acid Alone Prevents Neural Tube Defects: Evidence from the China Study
RJ Berry,Z Li
DOI: https://doi.org/10.1097/00001648-200201000-00021
2002-01-01
Epidemiology
Abstract:To the Editor: In a recent commentary, the call by Turner et al. for a new randomized clinical trial of folic acid (FA) alone to prevent neural tube defects (NTDs) 1 is based in part on the argument that the reduced risk for NTDs with FA in our China cohort study may have reflected between-group differences in some risk factors, rather than a true protective effect of FA. Although personal characteristics of our two comparison groups differed somewhat, such differences do not explain the reduction in NTD risk associated with FA in that study. In our cohort study, the main difference between women who did and did not take FA pills was whether they registered for prenatal health care before the end of the first trimester of pregnancy. 2Figure 1 shows that among women who registered for prenatal health care before their LMP, more than 80% bought and took FA pills. In contrast, fewer than 5% of women who registered after the first trimester took FA pills. Women who did not take pills were mostly those who were pregnant with their second child or who were more than 3 months’ pregnant when they registered for prenatal health care. These women did not have the opportunity to take periconceptional FA. As displayed in Figure 1, most of the women who took FA registered for prenatal care before their LMP, while most of those who did not take FA registered after their LMP. We found no evidence to suggest that cost barriers to pill-taking had any relation with not taking pills, as suggested by Turner et al. FIGURE 1: Percent of women registering for prenatal care who took folic acid and number of women registering for prenatal care who did and did not take folic acid, by week of registration in relation to last menstrual period (LMP).Because we knew that women in the north had higher rates of NTDs and that women in the north and south might differ in potential confounders such as education, occupation, and diet, we stratified all analyses by region. More importantly, within each region, we found negligible differences in educational level and occupation between women who did and did not take FA. We stated clearly in our report that adjustment for education and occupation did not change our results. We would like to provide further evidence for the lack of confounding by education and occupation among the women in our study who experienced the greatest reduction in risk for NTDs, ie, those with the periconceptional pattern of FA use. Table 1 shows the median percentage compliance with pill-taking (total number of pills taken divided by the total number of days between the starting and stopping dates) and the percentage of women who completed the recommended course and took folic acid until the end of the first trimester of pregnancy, by educational and occupational strata. Across these strata, daily compliance and completion of the course of pill taking were high and varied little. Table 1: Median Pill-Taking Compliance and Taking Pills Through First Trimester of Pregnancy among Women Who Took Periconceptional Folic Acid by Education, Occupation and Region, People’s Republic of China, October 1, 1993–December 31, 1996We would like to clarify some misunderstandings evident in the commentary by Turner et al. They state that our study included “women to whom supplements were not offered.” We described a campaign to prevent NTDs in China when the health authorities encouraged every woman planning to become pregnant to purchase and to take FA pills. Given the knowledge that FA prevents NTDs, not offering or refusing to allow women to purchase pills would have been unethical. During the final editing of our manuscript, our description of informed consent was changed to, “all women who took pills provided oral informed consent.” It should have read, “all women who had blood drawn provided oral informed consent.” Because all women were encouraged to take pills as part of a public health campaign to prevent NTDs, no informed consent was necessary. Turner et al. also state “women who reported having taken supplements containing only folic acid,” which implies that pill-taking was measured by questioning women about their use of FA pills. Women in this FA campaign used bottles, each containing 31 pills. At the end of each month a village health worker collected the bottles in her village, counted the pills left in the bottle, and submitted a monthly report for each woman to the county offices. This prospective, monthly collection of FA pill use for every woman, before outcomes were known, is a major strength of our evaluation. This information allowed us to calculate compliance and to examine woman-specific FA use before and after her LMP. In the accompanying editorial that discusses the strengths and limitations of the China study, 3 Moore concludes that the high NTD prevalence in the north can be reduced almost to the lowest NTD prevalence seen in the south by highly compliant use of FA during the periconceptional period (0.7 and 0.6/1,000 pregnancies of ≥20 weeks’ gestation, in the north and south respectively). We agree with Moore’s conclusion and believe that the China study provides solid evidence that daily use of 400 μg of FA without other vitamins prevents NTDs. Robert J. Berry Zhu Li