Maternal Periconceptional Vitamins: Interactions with Selected Factors and Congenital Anomalies?

Abstract

Background.  

The mechanisms by which folic acid may contribute to reductions in risk of several congenital anomalies are unknown. The data gap includes a lack of information on possible effect modification between maternal folic acid use and other maternal exposures. We hypothesized that effects of congenital anomalies associated with maternal fever, cigarette smoking or alcohol use would be modified by intake of vitamins.

Methods.  

We explored case-control data that showed risk reductions among infants and fetuses whose mothers consumed vitamins. Data were from California deliveries of infants and fetuses in the period 1987–1989. Maternal telephone interviews were completed for 207 (87%) conotruncal cases, 489 (85%) orofacial cleft cases, 265 (84%) neural tube defect cases, 165 (82%) limb anomaly cases, and 734 controls (nonmalformed infants).

Results.  

Considering women who reported vitamin use and no periconceptional fever as referents, for each anomaly group we observed elevated effects for the combinations of maternal vitamin use/fever, no use/no fever and no use/fever. Effects were most elevated for the combination of no vitamin use and fever. Adjusted for maternal body mass index, education and race/ethnicity, odds ratios were 2.4 (95% confidence inter-val = 1.0–5.9) for conotruncal defects, 2.9 (1.4–5.8) for cleft lip with or without cleft palate, 1.3 (0.4–3.9) for cleft palate, 3.1 (1.4–6.8) for neural tube defects, and 2.6 (1.0–6.4) for limb-deficiency defects. These interactions were further investigated relative to maternal use of fever-reducing medications. Effects tended to be highest among those women who did not use vitamins, had fevers, and did not use fever-reducing medications. Compared with women who used vitamins and did not smoke periconceptionally, anomaly risks tended to be highest among women who did not use vitamins and smoked. No specific pattern emerged involving alcohol intake.

Conclusions.  

These data further suggest that the underlying mechanisms of folic acid associated with congenital anomalies may be complex.

Over the last 2 decades, substantial epidemiologic evidence has emerged to indicate that periconceptional vitamin supplementation with folic acid reduces the risk of women having pregnancies affected by neural tube defects (NTDs) ^1,2 and perhaps other congenital anomalies. ^3–6 The underlying mechanisms by which folic acid reduces these risks, however, are unknown.

The current data gap includes a lack of information about factors that might modify the protective effects of maternal folic acid use. Hernandez-Diaz et al. ⁷ demonstrated that risks of selected congenital anomalies were increased when women used folate-antagonist medications during early pregnancy. These risks were attenuated if women also consumed a vitamin supplement containing folic acid during the same pregnancy time period. Botto et al. ⁸ recently demonstrated an association between maternal fever and selected heart anomalies. These investigators observed an attenuation in the risk associated with maternal fever among women who also consumed vitamins containing folic acid.

The current investigation extends observations made by Botto et al. ⁸ We hypothesized that the risks of selected congenital anomalies associated with particular factors (ie, maternal fever, maternal cigarette smoking and maternal alcohol consumption) would be modified in the presence of maternal periconceptional intake of vitamins containing folic acid. Investigating such relations between vitamin use and other risk factors could contribute to a more detailed understanding of the circumstances in which periconceptional vitamin use exerts its potential protective effects. In this study, we explored population-based case-control data that have previously demonstrated risk reductions among infants whose mothers used vitamin supplements containing folic acid around the time of conception.

Methods

Details of the population-based case-control data used in these analyses have been described. ^4,9 This study included NTDs, conotruncal heart defects, limb anomalies and orofacial clefts diagnosed within 1 year after birth among infants and fetal deaths delivered to women residing in most California counties. This included all deliveries of infants or fetal deaths (≥20 weeks’ gestation) between January 1987 and December 1988 (N = 344,214), except for orofacial cleft cases, which included deliveries through December 1989 (an additional 208,387 infants or fetal deaths considered eligible). Medical geneticists determined case eligibility by using detailed diagnostic information collected from medical records.

NTD cases comprised infants or fetuses (including elective terminations) with diagnoses of anencephaly, spina bifida cystica, craniorachischisis and iniencephaly confirmed by autopsy, surgery report, ultrasound or x-ray scan. Conotruncal heart cases comprised all infants and fetuses with anomalies affecting aorticopulmonary septation (includes tetralogy of Fallot, d-transposition of the great arteries, truncus arteriosus communis, double outlet right ventricle, pulmonary valve atresia with ventricular septal defect, subaortic ventricular septal defect type I and aorticopulmonary window) confirmed by echocardiography, cardiac catheterization, surgery or autopsy. Limb anomaly cases comprised all infants and fetuses with longitudinal, transverse or amniotic-band limb-deficiency defects of the upper or lower limbs confirmed by radiology, surgery or autopsy. Orofacial cleft cases comprised those infants or fetuses with cleft palate (CP) or with cleft lip with or without cleft palate (CLP) confirmed by surgery or autopsy. CP and CLP cases were further classified based on the nature of accompanying congenital anomalies. Cases with no other anomaly or with anomalies considered minor were classified as isolated. Cases with at least one accompanying major anomaly were classified as multiple. Only isolated CP and CLP cases were considered in these analyses. Infants diagnosed with single gene disorders, trisomies, or Turner syndrome (45,X) were excluded.

There were 972 control infants randomly selected from all infants born alive in the same geographic area and time period as cases (1987–1989). Of these controls, 652 were selected from the 1987–1988 birth cohort. Control infants had no major congenital anomalies identified before the first birthday. Interviews were completed with 265 (84%) NTD case mothers, 207 (87%) conotruncal case mothers, 165 (82%) limb anomaly case mothers, 489 (85%) orofacial cleft case mothers, and 734 (76%, [481 or 76% for the 1987–1988 cohort]) control mothers. Among the controls from the 1987–1988 cohort, interviews were conducted for 481 (76%) mothers. Interviews were completed within an average of 3.7 years from the date of delivery for cases, and 3.8 years for controls.

The protocol for this study was reviewed and approved by an institutional review board. All study participants provided their verbal consent to participate in interviews. Telephone interviews with case and control mothers elicited information on medical and reproductive histories, and activities associated with various lifestyles. The interviewer assisted each woman in establishing a 4-month periconceptional period (from 1 month before to 3 months after conception) that was referred to throughout the interview. Women were asked whether they used vitamin or mineral supplements during the 4-month period. Women who responded affirmatively were asked about the types of supplements used, and for each supplement the frequency of use per week and the quantity taken each time. With respect to fever, women were asked: “. . .during the 4-month period, did you have a fever of 100°F (37.8°C) or more?” If the woman responded affirmatively she was further asked, “Did you take medicines or receive injections for this condition? What were these?” To assess active maternal smoke exposures, women were asked how many cigarettes they smoked daily during the 4-month periconceptional period. To assess maternal alcohol consumption, women were asked how often they had a beverage containing alcohol in the 4-month period. In our analyses, we grouped maternal smoking or maternal alcohol consumption as any use vs no use.

We described the binary outcome of anomaly risk using additive linear logistic regression models. Effect estimation was limited to comparisons with 10 or more “exposed” case and control mothers combined. We considered as covariates maternal race/ethnicity (Latina, foreign-born; Latina, U.S.-born; white, non-Hispanic; black; other), education (<high school graduate; high school graduate; some college or college graduate), and body mass index (≤29kg/m²; >29kg/m²).

Results

Table 1 shows the associations of maternal periconceptional vitamin intake and maternal fever with selected congenital anomaly. Women who reported periconceptional vitamin use and no periconceptional fever were considered as referents. For each anomaly group, we observed modestly elevated effects for the combinations of maternal vitamin use/fever, no use/no fever and no use/fever. The associations were strongest for the combination of no vitamin use and fever, except for the isolated cleft palate group. Effect estimates in the s were adjusted for maternal body mass index, education and race/ethnicity, but adjustment had little effect.

TABLE 1: Association of Maternal Periconceptional Vitamin Intake and Maternal Fever with Selected Congenital Anomalies, Among Selected 1987–1989 California Deliveries

Because cereal is a source of folic acid, we explored whether frequent cereal use (more than once per week) by women in the periconceptional period further influenced the observed effects between fever and anomaly phenotype. Compared with referent women who did not have fever, used vitamin supplements, and were frequent consumers of cereal in the periconceptional period, the associations were strongest among those women who reported fever but did not use vitamin supplements or cereal (data not shown). However, data for these comparisons were sparse; only for isolated CLP were there 10 or more cases and controls available for analysis. The odds ratio (OR) for this grouping was 13 (95% confidence interval [CI] = 2.7–63).

We further investigated effects associated with maternal fever and vitamin use relative to maternal use of fever-reducing medications. We estimated the effects associated with the various possible combinations of maternal vitamins, fever and fever-reducing medication relative to the referent group of mothers who used vitamins and did not have a fever in the periconceptional period. Data were sufficient to estimate risks for only two anomaly groups (isolated CLP and NTDs)—for the combination of no vitamin use, fever and no fever-reducing medication use. For these two anomalies, effects tended to be highest (although imprecise) among those women who did not use vitamins, had fevers, and did not use fever-reducing medications. For isolated CLP, the odds ratio was 3.6 (CI = 1.0–12) and for NTDs the odds ratio was 7.4 (1.9–28).

We also examined whether effects of two other maternal exposures, specifically cigarette smoking and alcohol intake, might be influenced by the presence of maternal use of vitamin supplements. Compared with women who used vitamins and did not smoke periconceptionally, anomaly risks were highest among women who did not use vitamins and smoked, except for NTDs (Table 2). No specific pattern emerged involving periconceptional alcohol intake (Table 3).

TABLE 2: Association of Maternal Periconceptional Cigarette Smoking and Maternal Fever with Selected Congenital Anomalies, Among Selected 1987–1989 California Deliveries

TABLE 3: Association of Maternal Periconceptional Alcohol Intake and Maternal Fever with Selected Congenital Anomalies, Among Selected 1987–1989 California Deliveries

Discussion

Multivitamin supplements containing folic acid may modify the associations between maternal periconceptional fever and selected congenital anomalies, and possibly the associations with maternal smoking as well. Numerous studies have investigated the separate influences of vitamin intake on selected anomalies, ^1–6,10,11 of maternal fever on selected anomalies, ^12–15 and of maternal cigarette smoking on selected anomalies. ^9,16,17 However, only one previous study has reported whether maternal intake of folic acid modifies the effects of another risk factor. Botto and colleagues ⁸ observed that the association between maternal fever and offspring with specific heart anomalies was modified by maternal periconceptional vitamin use. For example, they observed the risk of heart defects associated with maternal fever to be 2.3 among infants whose mothers did not use vitamins and 1.1 among those infants whose mothers did use vitamin supplements. Our study extends their observation to a larger sample of specific heart anomalies and to additional anomalies whose etiologies have been observed to include maternal folic acid intakes. We also incorporated the influence of fever-reducing medication use and investigated the potential modification of effects of additional exposures (maternal cigarette and alcohol use).

We did not observe an interaction between maternal intake of multivitamins containing folic acid and maternal alcohol intake on the risk of congenital anomalies. This may be a function of no true modifying influence by vitamin use, or it may reflect the lack of an association between maternal alcohol and the studied anomalies. The literature reveals limited and rather inconsistent associations between maternal periconceptional alcohol intake and some of these anomalies. ^18–23

The underlying mechanisms by which folic acid may reduce risks of specific congenital anomalies are unknown. How folic acid may reduce risks of congenital anomalies in the presence of another potential teratogenic exposure such as maternal fever or maternal smoking is also unknown. Findings from experimental systems offer some support for our findings related to fever. One study observed that exposure of pregnant mice to both folic acid and hyperthermia resulted in a substantially lower prevalence of NTDs among offspring compared with exposure to hyperthermia alone. ²⁴ An earlier study, however, did not observe the same beneficial effects of folic acid on the prevalence of NTDs in the offspring of hyperthermia-treated golden hamsters. ²⁵

Several issues must be considered when interpreting our results. First, some of the risk factor/vitamin combinations yielded relatively small numbers of mothers, resulting in imprecise effect estimation. Thus, even though our findings are consistent with the interpretation that maternal vitamin use modified the effects of another exposure (fever or cigarette smoking), the imprecision associated with some effect estimates was large, leading us to question whether effect estimates were truly different across exposure-vitamin combinations. A second issue to consider is the potential for recall bias. It is possible that the accuracy of reported use of vitamins, fever episodes, smoking or consumption of alcohol was different for case mothers. Thus, we cannot exclude the possibility that observed effects were attributable to biased reporting (such as case mothers misreporting that they had a fever, or control mothers misreporting that they did not have a fever). Elevated effects owing to underreporting among controls seems unlikely, given that the frequencies of fever, smoking and alcohol consumption among controls were similar to or greater than what others have observed. ^8,15,26 The extent to which case mothers may have overreported the combination of no vitamin use and fever is unknown.

Despite the noted limitations, these data represent an extension of the initial findings by Botto et al. ⁸ and further suggest that the underlying mechanism by which folic acid supplementation reduces risks of some congenital anomalies may be complex.