Abstract
Objective:
The quantity and quality of dietary intake among women of reproductive age has important public health implications for nutritional status during pregnancy. We described dietary intake during the year before pregnancy among a large, diverse group of US mothers.
Methods:
We examined data from 11 109 mothers who gave birth from 1997 through 2011 and participated in a population-based case-control study, the National Birth Defects Prevention Study, as controls (mothers who had babies without major birth defects). We examined whether subgroups of mothers at elevated risk of adverse pregnancy outcomes were more likely than their reference groups to have high dietary intake (>90th percentile of intake) or low dietary intake (<10th percentile of intake). We examined dietary intake of 22 nutritional factors, which were estimated from responses to a food frequency questionnaire.
Results:
Participants who were aged <20, were nulliparous, had <high school diploma or <$20 000 annual household income, were non-Hispanic black, were underweight or obese, did not intend to become pregnant, did not take folic acid–containing vitamin supplements, or smoked had worse dietary intakes than their reference groups. For example, 17.5% of participants aged <20 had a low score on the diet quality index and 5.3% had a high score (vs expected values of 10%). Participants who were aged ≥35, were Hispanic, or had prepregnancy diabetes tended to have better dietary intakes than their reference groups. Maternal overweight and prepregnancy hypertension had few significant associations.
Conclusions:
Strategies are needed to ensure optimal nutrition among all childbearing women.
Keywords: dietary intake, preconception, elevated risk, adverse pregnancy outcomes, nutritional status
The quantity and quality of dietary intake among women of reproductive age has important public health implications because these factors set the stage for nutritional status during pregnancy. Nutritional status early in pregnancy (ie, the first several weeks) is important because fetal structures and the placenta develop during this period. Early pregnancy nutritional status is also associated with adverse maternal and infant health outcomes that occur later in pregnancy, such as gestational diabetes, pre-eclampsia, preterm delivery, and low birth weight.1-8
Almost half of all US pregnancies are unintended,9 and regardless of intention, women may not know they are pregnant until several weeks or more after conception. As a result, even if women do consciously decide to alter their diet in response to pregnancy, this change may not occur until many weeks into the pregnancy. Thus, good nutritional status should be reached before conception. Multiple consensus statements support the importance of maternal nutrition during pregnancy to maternal health and fetal development.10,11 However, the knowledge base that describes diet and its determinants before conception or during the earliest weeks of pregnancy is limited.12,13 Thus, a better understanding of diet before conception is needed so that strategies can be developed to improve the nutritional status of women during their childbearing years, which will ultimately improve maternal and infant health.
We examined dietary intake in the year before pregnancy as reported by a large, diverse, population-based group of mothers who gave birth in the United States from 1997 through 2011. These women participated as controls in the National Birth Defects Prevention Study (NBDPS). We examined whether subgroups of mothers considered to be at elevated risk of adverse pregnancy outcomes were more likely than mothers not at risk to have a high intake or a low intake of 22 selected nutritional factors.
Methods
Study Design
The NBDPS is a multistate, population-based case-control study that included deliveries with estimated due dates from 1997 through 2011. We restricted our analysis to mothers of infants who served as controls. The NBDPS was conducted according to the guidelines of the Declaration of Helsinki, and all procedures involving human participants were approved by the institutional review boards of the participating study centers and the Centers for Disease Control and Prevention. Verbal informed consent was obtained from all participants by telephone.
Detailed study methods and descriptions of surveillance systems in the 10 centers that contributed data to our analysis (Arkansas, California, Georgia, Iowa, Massachusetts, New Jersey, New York, North Carolina, Texas, and Utah) are available elsewhere.14 In brief, for controls, each participating center randomly selected at least 100 liveborn infants without major birth defects per study year from birth certificates or birth hospitals to represent the population from which the control participants were derived. Maternal interviews were conducted using a standardized, computer-based questionnaire, primarily by telephone, in English or Spanish, from 6 weeks to 24 months after the infant’s estimated date of delivery. Exposures to many factors were assessed, relative to the woman’s estimated date of conception, which was derived by subtracting 266 days from the estimated date of delivery. Interviews were conducted with mothers of 64% (11 829 of 18 614) of controls. Median time from actual date of delivery to interview was 7.6 months (interquartile range, 5.0-11.7 months), so that, on average, the year before pregnancy spanned approximately 17 to 29 months from time of interview.
Dietary Assessment and Nutritional Factors
Participants reported their average intake of foods during the year before they became pregnant by answering a 58-item food frequency questionnaire developed for The Nurses’ Health Study.15 Intake of breakfast cereals, sodas, and food supplements (eg, powdered drink supplements) was assessed by separate, more detailed questions, which covered intake during the 3 months before pregnancy. Because few participants (<10%) consumed food supplements and nutrient data were not available for many of these products, we did not include food supplements in nutrient calculations. We used the US Department of Agriculture’s nutrient database version 27 as the source of nutrient values.16 Beginning with deliveries in 2006, several food items were added to the food frequency questionnaire; to ensure comparability of data during the study, we did not consider these additional food items in nutrient calculations. Of the 11 829 participants interviewed, we excluded 720 who reported a daily energy intake of <500 kcal or >5000 kcal or who had >1 missing food item from the food frequency questionnaire, leaving 11 109 participants for analysis.
We selected the following nutritional factors for study on the basis of their emphasis in dietary guidelines for pregnant women and previous associations with adverse pregnancy outcomes: beta carotene, calcium, choline, folate, iron, lycopene, niacin, retinol, riboflavin, thiamin, vitamins A and C, and servings per day of fruits, vegetables, dairy, grains, and sweets. We also calculated the percentage of calories from protein, fat, and saturated fat; dietary glycemic index; and a diet quality index (DQI). The dietary glycemic level was calculated by summing the glycemic index of each food, multiplied by average daily servings and carbohydrate content of each food, and dividing the sum by total average daily carbohydrate intake; the glycemic index reflects the quality of dietary carbohydrate intake, relative to the blood glucose response to consumption of a food item. The DQI examines intake of nutrients and food groups related to nutritional recommendations during pregnancy; it is based on a previously validated index17,18 that we adapted to the NBDPS food frequency questionnaire.19 The DQI is the summary score of 6 positively scored components (grains, vegetables, fruits, folate, iron, and calcium) and 2 negatively scored components (intake of sweets and percentage of calories from fat). For each participant, we scored each component from 0 to 3 on the basis of quartiles of the distribution among controls, and then we summed the component scores to obtain the final value for the DQI; the score could range from 0 to 18, and a higher score indicated better diet quality.
We standardized nutritional factors to the median energy intake among all study participants: for each participant, we divided the value of each nutritional factor by her average daily energy intake and then multiplied it by the median energy intake among all participants (1461.32 kcal/d). We did this for each nutritional factor except glycemic index and percentage of calories from protein, fat, and saturated fat, because they are inherently independent of energy intake.
For most studied nutritional factors, we considered a higher intake to be “better” and a lower intake “worse.” Exceptions were that lower intake was considered better for servings per day of sweets, a lower glycemic index was considered better, and a lower percentage of calories from fat and saturated fat was considered better.
Maternal Characteristics
We examined dietary intake among participants with selected characteristics that are associated with adverse maternal and infant health outcomes, hereinafter referred to as “risk groups.” The risk groups and reference groups were participants aged <20 and participants aged ≥35, compared with participants aged 20-34; nulliparity (no previous live births) and high parity (≥3 previous live births), compared with 1 or 2 live births previous to the index delivery; low level of education (<high school diploma), compared with ≥high school diploma; low annual household income (≤$20 000), compared with higher income (>$20 000); non-Hispanic black or Hispanic race/ethnicity, each compared with non-Hispanic white; unintended pregnancy, compared with intended pregnancy (based on mother’s self-report); obese, overweight, or underweight body mass index (BMI) before pregnancy, compared with normal BMI (BMI calculated as self-reported prepregnancy weight in kilograms divided by height in meters squared [kg/m2]; obesity was defined as BMI ≥30.0, overweight as 25.0-29.9, normal weight as 18.5-24.9, and underweight as <18.5 kg/m2)20; no intake of folic acid–containing vitamin supplements during the 3 months before pregnancy, compared with any intake; smoking in the month before pregnancy, compared with no smoking; prepregnancy diabetes, compared with no diabetes; and prepregnancy hypertension, compared with no prepregnancy hypertension.
Analyses
For each nutritional factor, we created a variable indicating whether a participant’s intake was low (defined as <10th percentile) or high (defined as >90th percentile) relative to the reference group (10th-90th percentile) on the basis of cutoffs that were derived from the distribution among all study participants (Table 1). For each risk group and nutritional factor, we examined whether the percentage of participants in each of these percentile categories was different from the percentage in the reference group. We used Pearson χ2 tests to determine differences and considered P < .05 to be significant. Under the null hypothesis, about 10% of participants in each category of a risk factor, including the risk group, would have high intake and 10% would have low intake; for simplicity of interpretation, we referred to this as the “expected” percentage. For convenience, we categorized intake as low or high, although we acknowledge that these categories are relevant to our study sample rather than to a particular dietary recommendation, because this study was based on uncalibrated data from a food frequency questionnaire. In addition, to focus the study’s results on intake among risk groups, we tabulated data on intake among risk groups only; we did not tabulate data on intake among participants in the reference groups. Analyses for each risk factor included all participants who had nonmissing data on that factor. We used SAS version 9.1 for all analyses.21
Table 1.
Percentile cutoffs for average daily intake for nutritional factors among control mothers who participated in the National Birth Defects Prevention Study, 1997-2011a
| Nutritional Factor | 10th Percentile | 90th Percentile |
|---|---|---|
| Diet quality indexb | 5.9 | 17.6 |
| Nutrient | ||
| Beta carotene, µg | 623.5 | 5526.5 |
| Calcium, mg | 407.1 | 1151.4 |
| Choline, mg | 206.4 | 399.2 |
| Folate, µg dietary folate equivalent | 253.0 | 776.6 |
| Iron, mg | 7.0 | 19.6 |
| Lycopene, µg | 0.3 | 8128.8 |
| Niacin, mg | 13.0 | 26.4 |
| Retinol, µg | 192.2 | 755.6 |
| Riboflavin, mg | 1.2 | 2.6 |
| Thiamin, mg | 0.8 | 1.6 |
| Vitamin A, µg retinol activity equivalent | 358.2 | 1119.1 |
| Vitamin C, mg | 41.5 | 189.4 |
| No. of servings per day | ||
| Fruits | 0.6 | 3.8 |
| Vegetables | 0.8 | 3.6 |
| Dairy | 0.3 | 3.6 |
| Grains | 1.0 | 4.0 |
| Sweets | 0.3 | 3.3 |
| Dietary glycemic indexc | 46.2 | 58.3 |
| Percentage of calories from… | ||
| Protein | 13.2 | 23.0 |
| Fat | 21.0 | 37.1 |
| Saturated fat | 7.5 | 14.8 |
a Data source: Yoon et al.14 Nutritional factors were standardized to the median energy intake among all 11 109 study participants (1461.32 kcal/d) before determining percentile cutoffs. Values are from food frequency questionnaire data and therefore should be interpreted as semiquantitative and not as absolutes.
b A summary score of 6 positively scored components (grains, vegetables, fruits, folate, iron, and calcium) and 2 negatively scored components (intake of sweets and percentage of calories from fat). For each participant, we scored each component from 0 to 3 on the basis of quartiles of the distribution among controls, and then we summed the component scores to obtain the final value for the diet quality index; the score could range from 0 to 18, and a higher score indicates better diet quality. Index is based on a previously validated index17,18 adapted to the National Birth Defects Prevention Study food frequency questionnaire.19
c Calculated by summing the glycemic index of each food, multiplied by average daily servings and carbohydrate content of each food, and dividing the sum by total average daily carbohydrate intake; the glycemic index reflects the quality of dietary carbohydrate intake, relative to the blood glucose response to consumption of a food item.
Results
Of the 11 109 study participants, most were aged 20-34 (76.3%; n = 8471), had ≥high school diploma (83.1%; n = 9230), had normal BMI (51.4%; n = 5708), and were non-Hispanic white (59.0%; n = 6551; Table 2). Except for the risk group defined as nonusers of vitamins or minerals (63.4%; n = 7042), risk groups comprised a minority of participants.
Table 2.
Characteristics of study participants and designation of risk groupsa and reference groupsb for comparisons of distributions of nutritional factors, among control mothers who participated in the National Birth Defects Prevention Study, 1997-2011c
| Characteristic | No. (%d) (n = 11 109) |
|---|---|
| Age, y | |
| <20a | 1073 (9.7) |
| 20-24b | 2490 (22.4) |
| 25-29b | 3099 (27.9) |
| 30-34b | 2882 (25.9) |
| ≥35a | 1565 (14.1) |
| Parity (no. of previous live births) | |
| 0a | 4398 (39.6) |
| 1b | 3633 (32.7) |
| 2b | 1914 (17.2) |
| ≥3a | 1161 (10.5) |
| Missing data | 3 (<0.1) |
| Education | |
| <High school diplomaa | 1798 (16.2) |
| High school diplomab | 2617 (23.6) |
| Some collegeb | 2982 (26.8) |
| College graduateb | 3631 (32.7) |
| Missing data | 81 (0.7) |
| Annual household income, $ | |
| ≤20 000a | 3212 (28.9) |
| >20 000b | 7039 (63.4) |
| Missing data | 858 (7.7) |
| Race/ethnicity | |
| Non-Hispanic whiteb | 6551 (59.0) |
| Non-Hispanic blacka | 1191 (10.7) |
| Hispanica | 2645 (23.8) |
| Other non-Hispanic race/ethnicity | 718 (6.5) |
| Missing data | 4 (<0.1) |
| Pregnancy intention | |
| Intended to be pregnantb | 6749 (60.8) |
| Wanted to wait until later | 1818 (16.4) |
| Did not intend to become pregnanta | 1036 (9.3) |
| Did not care | 792 (7.1) |
| Missing data | 714 (6.4) |
| Prepregnancy body mass indexe | |
| Underweighta | 568 (5.1) |
| Normal weightb | 5708 (51.4) |
| Overweighta | 2437 (21.9) |
| Obesea | 1947 (17.5) |
| Missing data | 449 (4.0) |
| Took folic acid–containing vitamin supplements 3 months before pregnancy | |
| Noa | 7042 (63.4) |
| Yesb | 3934 (35.4) |
| Missing data | 133 (1.2) |
| Smoked cigarettes 1 month before pregnancy | |
| Nob | 9106 (82.0) |
| Yesa | 1965 (17.7) |
| Missing data | 38 (0.3) |
| Prepregnancy diabetes | |
| Nob | 10 830 (97.5) |
| Yesa | 263 (2.4) |
| Missing data | 16 (0.1) |
| Prepregnancy hypertension | |
| Nob | 10 592 (95.4) |
| Yesa | 500 (4.5) |
| Missing data | 17 (0.2) |
a Risk groups were defined on the basis of selected characteristics associated with adverse maternal and infant health outcomes. Each risk group was compared separately with its reference group in analyses.
b Reference groups were compared with risk groups in analyses; for categories with >1 reference group, reference groups were combined into a single category.
c Data source: Yoon et al.14
d Percentages may not add to 100 because of rounding.
e Body mass index (BMI) calculated as self-reported prepregnancy weight in kilograms divided by height in meters squared (kg/m2); obesity was defined as BMI ≥30.0, overweight as 25.0-29.9, normal weight as 18.5-24.9, and underweight as <18.5 kg/m2.20
Participants aged <20 were significantly more likely than participants aged 20-34 to have low intake of most nutritional factors (Table 3). For example, 17.5% of participants aged <20 had a low score on the DQI; this 17.5% was significantly lower than the expected value of 10%. We found fewer significant differences in high intake of nutritional factors. In contrast, participants aged ≥35 were significantly less likely than participants aged 20-34 to have low intake for most nutritional factors; for example, 6.0% of participants aged ≥35 had a low score on the DQI. Participants who were nulliparous (before the index pregnancy) were significantly more likely than participants with 1 or 2 previous live births to have low intake for most nutritional factors, but the differences from the expected 10% tended to be modest (ie, <2 percentage-point difference); for example, 11.3% of nulliparous participants had a low score on the DQI. Participants who had ≥3 previous live births were less likely than participants with 1 or 2 previous live births to have low intake and more likely to have high intake of several nutritional factors.
Table 3.
Percentage of study participants in selected risk groupsa who had low dietary intake or high dietary intakeb of selected nutritional factors, by maternal age group, parity, folic acid–containing vitamin supplement intake, and smoking status, and whether these percentages were significantly different from the reference groups, among control mothers who participated in the National Birth Defects Prevention Study, 1997-2011c
| Age Groupd | Paritye | No Folic Acid–Containing Vitamin Supplement Intake During 3 Months Before Pregnancyf | Smoking During 1 Month Before Pregnancyg | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Aged <20 | Aged ≥35 | Nulliparous | ≥3 Previous Live Births | |||||||||
| Factor | % Low | % High | % Low | % High | % Low | % High | % Low | % High | % Low | % High | % Low | % High |
| Positively scored factors | ||||||||||||
| Diet quality indexh | 17.5i | 5.3i | 6.0j | 15.0j | 11.3i | 10.8j | 8.4 | 8.2 | 11.5i | 9.0i | 17.9i | 5.9i |
| Nutrients | ||||||||||||
| Beta carotene | 23.3i | 5.0i | 2.5j | 16.4j | 11.7i | 9.3 | 7.1j | 10.7 | 12.7i | 9.3 | 17.0i | 5.2i |
| Calcium | 19.9i | 5.8i | 6.2j | 12.5j | 11.2i | 10.4 | 9.0 | 9.2 | 12.1i | 8.0i | 17.2i | 8.1 |
| Choline | 20.6i | 8.9 | 4.3j | 11.4 | 12.0i | 10.0 | 7.1 | 13.4j | 11.8i | 10.6j | 17.5i | 9.5 |
| Folate | 17.2i | 9.3 | 5.5j | 10.7 | 11.6i | 10.6j | 8.6 | 8.9 | 11.5i | 9.4 | 17.7i | 9.7 |
| Iron | 16.2i | 10.0 | 9.2 | 10.0 | 11.5i | 10.3 | 10.0 | 9.0 | 10.7i | 9.8 | 15.0i | 10.5 |
| Lycopene | 22.6i | 10.4j | 4.3j | 8.9i | 11.3i | 9.6 | 8.4 | 13.7j | 12.4i | 11.3j | 13.7i | 7.6i |
| Niacin | 20.2i | 7.5 | 5.2j | 10.7 | 11.1i | 11.1j | 9.9 | 8.4 | 12.1i | 9.6 | 14.9i | 10.3 |
| Retinol | 15.3i | 10.3 | 7.6j | 8.9 | 11.6i | 9.4 | 8.8 | 13.4j | 11.2i | 11.0j | 12.4i | 9.0 |
| Riboflavin | 20.6i | 6.2i | 3.9j | 13.9j | 12.0i | 10.6 | 7.7 | 8.3 | 12.1i | 8.9i | 16.9i | 12.1j |
| Thiamin | 17.4i | 9.8 | 4.7j | 10.5 | 11.0i | 10.4 | 8.2 | 9.0 | 11.7i | 9.6 | 18.8i | 8.8 |
| Vitamin A | 22.1i | 7.5 | 4.2j | 12.1j | 11.8i | 8.8i | 7.1 | 13.4j | 12.3i | 10.7j | 15.8i | 6.4i |
| Vitamin C | 13.3i | 12.2j | 4.9j | 11.5 | 10.3 | 9.7 | 6.9j | 12.8j | 10.9i | 11.6j | 18.7i | 6.1i |
| Fruits | 13.9i | 12.1j | 6.3j | 11.9j | 10.8 | 9.6 | 7.8 | 12.7j | 11.4i | 10.7j | 21.7i | 5.0i |
| Vegetables | 24.8i | 8.0 | 4.5j | 13.5j | 11.6i | 9.8 | 7.3 | 14.4j | 12.3i | 10.3j | 14.6i | 7.7i |
| Dairy | 16.2i | 6.4i | 7.6j | 10.8 | 11.3i | 9.8 | 8.8 | 10.2 | 11.3i | 8.3i | 15.6i | 8.3 |
| Grains | 16.4i | 10.9j | 8.6 | 10.0 | 11.3i | 10.2 | 9.5 | 8.8 | 11.3i | 10.1 | 15.3i | 8.8 |
| % Calories from protein | 24.5i | 4.9i | 3.6j | 12.4 | 11.7i | 10.7j | 8.6 | 8.9 | 12.9i | 9.2i | 17.6i | 10.4j |
| Negatively scored factors | ||||||||||||
| Sweets | 5.4i | 21.2i | 14.3j | 3.5j | 9.7 | 10.9i | 11.4 | 8.5 | 8.9i | 12.6i | 9.4 | 21.6i |
| Glycemic index | 6.1i | 19.6i | 14.7j | 3.1j | 9.5 | 12.0i | 12.5j | 8.0 | 9.1i | 12.3i | 5.5i | 21.8i |
| % Calories from fat | 14.3j | 8.1 | 7.7i | 11.0 | 10.2 | 9.2 | 9.3 | 11.0 | 11.6j | 9.6 | 11.1j | 13.4i |
| % Calories from saturated fat | 14.5j | 8.3 | 7.2i | 10.0 | 10.5 | 9.5 | 9.2 | 9.6 | 11.7j | 9.2j | 10.5 | 11.7i |
a Risk groups were defined on the basis of selected characteristics associated with adverse maternal and infant health outcomes.
b High intake refers to dietary intake >90th percentile, and low intake refers to dietary intake <10th percentile; percentiles were based on the distribution of dietary intake among all study participants (Table 1 presents percentile cutoffs).
c Data source: Yoon et al.14
d Reference group was participants aged 20-34.
e Reference group was participants with 1 or 2 previous live births.
f Reference group was participants who did take folic acid–containing vitamin supplements during the 3 months before pregnancy.
g Reference group was participants who did not smoke during the 3 months before pregnancy.
h A summary score of 6 positively scored components (grains, vegetables, fruits, folate, iron, and calcium) and 2 negatively scored components (intake of sweets and percentage of calories from fat). For each participant, we scored each component from 0 to 3 on the basis of quartiles of the distribution among controls, and then we summed the component scores to obtain the final value for the DQI; the score could range from 0 to 18, and a higher score indicates better diet quality. Index is based on a previously validated index17,18 adapted to the National Birth Defects Prevention Study food frequency questionnaire.19
i Risk group had significantly (P < .05) worse intake than reference group, as determined by Pearson χ2 tests. That is, for positively scored nutrients, a larger percentage of participants than expected had low intake (ie, >10% had intake <10th percentile) or a smaller percentage of participants than expected had high intake (ie, <10% had intake >90th percentile). For negatively scored nutrients, a smaller percentage of participants than expected had low intake (ie, <10% had intake <10th percentile) or a larger percentage of participants than expected had high intake (ie, >10% had intake >90th percentile).
j Risk group had significantly (P < .05) better intake than reference group, as determined by Pearson χ2 tests. For positively scored nutrients, a smaller percentage of participants than expected had low intake (ie, <10% had intake <10th percentile) or a larger percentage of participants than expected had high intake (ie, >10% had intake >90th percentile). For negatively scored nutrients, a larger percentage of participants than expected had low intake (ie, >10% had intake <10th percentile) or a smaller percentage of participants than expected had high intake (ie, <10% had intake >90th percentile).
Participants who did not take folic acid–containing vitamin supplements during the 3 months before pregnancy, compared with those who did, were more likely to have low intake of most nutritional factors. For example, 11.5% of women who took supplements had a low score on the DQI. We found significant differences in high intake for many factors, but these differences were both positive and negative and most were modest (<2 percentage-point difference). Participants who smoked had worse intake for most nutritional factors, and we found some of the most marked differences in the comparison of smokers and nonsmokers. For example, 17.9% of smokers had a low score on the DQI, and 5.9% had a high score.
Participants with <high school diploma, participants with ≤$20 000 annual household income, and participants who were non-Hispanic black had significantly worse intakes of most nutritional factors than their reference groups, with the exception of fats, for which they had better intake (ie, they were more likely to have low intake of fats; Table 4). However, these 3 groups were also more likely than their reference groups to have high intake of several nutritional factors (eg, vitamin C, retinol, and fruits).
Table 4.
Percentage of study participants in selected risk groupsa who had low dietary intake or high dietary intakeb of selected nutritional factors, by maternal education, household income, race/ethnicity, and pregnancy intention, and whether these percentages were significantly different from the reference groups, among control mothers who participated in the National Birth Defects Prevention Study, 1997–2011c
| Education <High School Diplomad | Annual Household Income ≤$20 000e | Race/Ethnicityf | Did Not Intend to Become Pregnantg | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Non-Hispanic Black | Hispanic | |||||||||
| Factor | % Low | % High | % Low | % High | % Low | % High | % Low | % High | % Low | % High |
| Positively scored factors | ||||||||||
| Diet quality indexh | 12.1i | 6.8i | 12.5i | 7.0i | 18.2i | 7.9 | 5.1j | 9.0i | 14.2i | 8.2i |
| Nutrients | ||||||||||
| Beta carotene | 15.4i | 8.7 | 14.9i | 8.1i | 17.3i | 11.6j | 10.0i | 11.4j | 15.0i | 9.5 |
| Calcium | 14.7i | 5.7i | 14.2i | 7.3i | 19.6i | 5.0i | 10.4i | 5.6i | 15.8i | 5.5i |
| Choline | 14.0i | 12.3j | 13.7i | 11.6j | 13.9i | 14.2j | 7.0j | 13.7j | 13.8i | 12.6j |
| Folate | 12.5i | 9.3 | 12.8i | 9.5 | 13.5i | 11.2 | 7.4j | 8.6i | 11.8i | 9.7 |
| Iron | 11.2i | 10.9 | 12.0i | 10.4 | 13.0i | 11.6j | 7.0j | 10.7 | 10.3 | 11.8j |
| Lycopene | 17.3i | 17.5j | 15.4i | 13.6j | 24.4i | 2.3i | 10.6i | 25.1j | 14.1i | 12.9j |
| Niacin | 18.9i | 7.6i | 16.5i | 8.1i | 10.8i | 14.1j | 13.0i | 7.4i | 11.4i | 9.7 |
| Retinol | 13.8i | 15.8j | 12.2i | 13.8j | 10.9i | 15.5j | 12.6i | 14.5j | 12.9i | 13.1j |
| Riboflavin | 14.3i | 7.7i | 14.4i | 7.7i | 16.3i | 8.2i | 9.1 | 6.5i | 12.4i | 9.1 |
| Thiamin | 13.6i | 10.0 | 13.7i | 9.7 | 13.6i | 13.9j | 7.8j | 9.1 | 12.6i | 10.7 |
| Vitamin A | 15.7i | 14.3j | 14.3i | 12.0j | 13.2i | 14.1j | 12.4i | 15.3j | 14.9i | 12.2j |
| Vitamin C | 9.6 | 16.8i | 11.1i | 13.6j | 10.7 | 14.0j | 4.0j | 17.7j | 12.0i | 12.1j |
| Fruits | 10.5 | 15.1j | 11.6i | 12.3j | 13.4i | 12.8j | 4.3j | 16.9j | 13.6i | 10.6 |
| Vegetables | 16.7i | 13.2j | 15.9i | 10.6j | 19.4i | 6.0 | 11.9i | 17.0j | 15.2i | 11.5j |
| Dairy | 13.0i | 6.5i | 13.1i | 7.8i | 18.2i | 4.6i | 9.5 | 6.0i | 14.0i | 6.6i |
| Grains | 13.2i | 10.7 | 12.6i | 10.5j | 14.4i | 13.5j | 9.5 | 10.0j | 11.9i | 12.2j |
| % Calories from protein | 18.1i | 5.5i | 16.2i | 6.9i | 16.8i | 11.3 | 11.1i | 6.4i | 13.9i | 9.7 |
| Negatively scored factors | ||||||||||
| Sweets | 7.8i | 15.5i | 7.8i | 15.3i | 8.6 | 16.1i | 9.6 | 8.2j | 9.0 | 14.7i |
| Glycemic index | 13.8j | 13.2i | 9.7 | 14.3i | 4.2i | 20.7i | 16.3j | 4.9j | 8.0i | 15.8i |
| % Calories from fat | 16.3j | 6.3j | 13.5j | 8.8j | 11.7j | 12.1 | 15.0j | 4.8j | 12.2j | 10.7 |
| % Calories from saturated fat | 16.7j | 6.0j | 14.5j | 8.5j | 13.4j | 10.0 | 16.1j | 3.5j | 13.1j | 10.2 |
a Risk groups were defined on the basis of selected characteristics associated with adverse maternal and infant health outcomes.
b High intake refers to dietary intake >90th percentile, and low intake refers to dietary intake <10th percentile; percentiles were based on the distribution of dietary intake among all study participants (Table 1 presents percentile cutoffs).
c Data source: Yoon et al.14
d Reference group was participants with a high school diploma or greater education.
e Reference group was participants with annual household income >$20 000.
f Reference group was non-Hispanic white.
g Reference group was participants who did intend to become pregnant.
h A summary score of 6 positively scored components (grains, vegetables, fruits, folate, iron, and calcium) and 2 negatively scored components (intake of sweets and percentage of calories from fat). For each participant, we scored each component from 0 to 3 on the basis of quartiles of the distribution among controls, and then we summed the component scores to obtain the final value for the diet quality index; the score could range from 0 to 18, and a higher score indicates better diet quality. Index is based on a previously validated index17,18 adapted to the National Birth Defects Prevention Study food frequency questionnaire.19
i Risk group had significantly (P < .05) worse intake than the reference group, as determined by Pearson χ2 tests. That is, for positively scored nutrients, a larger percentage of participants than expected had low intake (ie, >10% had intake <10th percentile) or a smaller percentage of participants than expected had high intake (ie, <10% had intake >90th percentile). For negatively scored nutrients, a smaller percentage of participants than expected had low intake (ie, <10% had intake <10th percentile) or a larger percentage of participants than expected had high intake (ie, >10% had intake >90th percentile).
j Risk group had significantly (P < .05) better intake than reference group, as determined by Pearson χ2 tests. For positively scored nutrients, a smaller percentage of participants than expected had low intake (ie, <10% had intake <10th percentile) or a larger percentage of participants than expected had high intake (ie, >10% had intake >90th percentile). For negatively scored nutrients, a larger percentage of participants than expected had low intake (ie, >10% had intake <10th percentile) or a smaller percentage of participants than expected had high intake (ie, <10% had intake >90th percentile).
Results were more mixed for Hispanic participants than for participants who were non-Hispanic black. We found significantly better intake and significantly worse intake for several nutritional factors among Hispanic participants compared with non-Hispanic white participants. For example, fewer Hispanic participants (5.1%) had a low score on the DQI, but fewer (9.0%) had a high score on the DQI as well, relative to non-Hispanic white participants.
Participants who did not intend to become pregnant were significantly more likely than participants who did intend to become pregnant to have low intake of most nutritional factors. For several factors, they were also significantly more likely to have high intake (calcium, vitamin C, retinol, vitamin A, total choline, lycopene, vegetables, and grains).
Underweight participants were significantly more likely than normal-weight participants to have low intake of most nutritional factors; for example, 12.3% of underweight participants had a low score on the DQI (Table 5). We found few significant differences for overweight participants. Obese participants were more likely than normal-weight participants to have low intake of several nutritional factors, including folate, and they were more likely to have a low score on the DQI and a high intake of sweets.
Table 5.
Percentage of study participants in selected risk groupsa who had low dietary intake or high dietary intake of selected nutritional factors, by maternal prepregnancy body mass index, diabetes, and hypertension, and whether these percentages were significantly different from the reference groups, among control mothers who participated in the National Birth Defects Prevention Study, 1997-2011b
| Prepregnancy Body Mass Indexc | Prepregnancy Diabetesd | Prepregnancy Hypertensione | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Underweight | Overweight | Obese | ||||||||
| Factor | % Low | % High | % Low | % High | % Low | % High | % Low | % High | % Low | % High |
| Positively scored factors | ||||||||||
| Diet quality indexf | 12.3g | 8.8 | 9.7 | 10.2 | 12.6g | 8.2g | 6.1h | 9.5 | 7.8 | 9.6 |
| Nutrients | ||||||||||
| Beta carotene | 14.8g | 9.9 | 9.2 | 9.4 | 10.9 | 8.6g | 8.7 | 9.9 | 8.2 | 9.2 |
| Calcium | 14.4g | 9.0 | 9.2 | 9.6g | 11.0 | 8.5g | 9.9 | 9.9 | 7.8 | 9.2 |
| Choline | 17.1g | 6.9 | 7.8h | 10.4 | 11.0 | 11.4h | 4.9h | 14.8h | 8.4 | 9.8 |
| Folate | 13.9g | 8.6 | 9.4 | 9.9 | 12.5g | 8.5g | 4.2h | 10.6 | 8.8 | 7.2g |
| Iron | 13.6g | 9.2 | 9.4 | 9.9 | 11.6g | 9.2 | 7.2 | 9.9 | 10.0 | 7.6 |
| Lycopene | 12.7g | 7.2 | 8.5 | 11.2h | 12.3g | 10.9h | 8.0 | 12.5 | 8.2 | 6.8g |
| Niacin | 14.8g | 9.2 | 8.5 | 10.8 | 11.0g | 10.5 | 4.6h | 11.0 | 6.4h | 8.6 |
| Retinol | 12.0 | 8.8 | 8.5 | 9.5 | 11.0 | 10.9h | 9.5 | 12.5 | 8.0 | 8.6 |
| Riboflavin | 12.7g | 7.2g | 8.6 | 9.8 | 13.7g | 8.6g | 6.1h | 11.8 | 6.4h | 10.0 |
| Thiamin | 13.2g | 9.9 | 8.5 | 10.8 | 13.7g | 7.8g | 6.1h | 8.7 | 11.2 | 7.2g |
| Vitamin A | 12.3g | 9.9 | 8.6 | 9.8 | 11.6g | 9.8 | 6.8 | 12.9 | 5.2h | 9.0 |
| Vitamin C | 13.0g | 10.2 | 9.4 | 10.6h | 11.6g | 8.9 | 6.5h | 6.8 | 10.8 | 7.0g |
| Fruits | 13.9g | 11.6h | 9.8 | 10.2 | 11.4 | 8.1 | 11.0 | 9.1 | 10.8 | 8.2 |
| Vegetables | 13.7g | 8.1 | 9.1 | 10.6h | 10.6 | 11.0h | 6.5 | 14.4h | 8.2 | 9.0 |
| Dairy | 12.9g | 8.6 | 9.6 | 10.1 | 11.0 | 8.8g | 8.7 | 11.4 | 10.0 | 10.0 |
| Grains | 10.2 | 12.3 | 9.9 | 10.3 | 11.7g | 7.8g | 10.3 | 10.6 | 9.8 | 8.4 |
| % Calories from protein | 16.5g | 7.0 | 8.7 | 10.4 | 11.1g | 12.0h | 5.7h | 16.0h | 7.4 | 12.0 |
| Negatively scored factors | ||||||||||
| Sweets | 8.5 | 13.4g | 10.1 | 8.7 | 8.7 | 14.2g | 16.7h | 6.5 | 11.0 | 9.4 |
| Glycemic index | 8.1 | 15.3g | 10.2 | 9.5 | 9.2 | 11.5g | 13.7h | 8.7 | 9.2 | 7.8 |
| % Calories from fat | 10.4 | 8.8 | 9.1 | 10.1 | 10.2 | 10.6 | 6.5 | 13.7 | 8.0 | 12.8g |
| % Calories from saturated fat | 10.7 | 9.5 | 8.5 | 9.8 | 10.3 | 10.5 | 7.2 | 12.2 | 8.4 | 11.0 |
a Risk groups were defined on the basis of selected characteristics associated with adverse maternal and infant health outcomes.
b Data source: Yoon et al.14
c Reference group was normal weight.
d Reference group was participants without diabetes.
e Reference group was participants without hypertension.
f A summary score of 6 positively scored components (grains, vegetables, fruits, folate, iron, and calcium) and 2 negatively scored components (intake of sweets and percentage of calories from fat). For each participant, we scored each component from 0 to 3 on the basis of quartiles of the distribution among controls, and then we summed the component scores to obtain the final value for the diet quality index; the score could range from 0 to 18, and a higher score indicates better diet quality. Index is based on a previously validated index17,18 adapted to the National Birth Defects Prevention Study food frequency questionnaire.19
g Risk group had significantly (P < .05) worse intake than reference group, as determined by Pearson χ2 tests. That is, for positively scored nutrients, a larger percentage of participants than expected had low intake (ie, >10% had intake <10th percentile) or a smaller percentage of participants than expected had high intake (ie, <10% had intake >90th percentile). For negatively scored nutrients, a smaller percentage of participants than expected had low intake (ie, <10% had intake <10th percentile) or a larger percentage of participants than expected had high intake (ie, >10% had intake >90th percentile).
h Risk group had significantly (P < .05) better intake than reference group, as determined by Pearson χ2 tests. For positively scored nutrients, a smaller percentage of participants than expected had low intake (ie, <10% had intake <10th percentile) or a larger percentage of participants than expected had high intake (ie, >10% had intake >90th percentile). For negatively scored nutrients, a larger percentage of participants than expected had low intake (ie, >10% had intake <10th percentile) or a smaller percentage of participants than expected had high intake (ie, <10% had intake >90th percentile).
Participants with prepregnancy diabetes were significantly less likely than participants without prepregnancy diabetes to have a low intake of several nutritional factors, such as folate (4.2%), have a lower score on the DQI (6.1%), and take in a smaller percentage of calories from protein (5.7%), and they were more likely to have a low intake of sweets (16.7%). We found few significant differences between participants with prepregnancy hypertension and participants without prepregnancy hypertension.
Discussion
This multicentered US study of more than 11 000 participants observed that population subgroups at elevated risk of adverse pregnancy outcomes had worse dietary intake during the year before they became pregnant, relative to their lower-risk comparator groups. The most marked differences were for participants aged <20, participants with low levels of education and annual household income, non-Hispanic black participants, underweight participants, participants whose pregnancy was unintended, and participants who smoked. Participants who had ≥3 previous live births, were aged ≥35, were Hispanic, or had prepregnancy diabetes tended to have better dietary intake, relative to their reference groups. Participants who were overweight or had prepregnancy hypertension differed significantly from their reference groups for only a few factors.
The literature on preconception dietary intake is limited, given the importance of maternal nutrition before pregnancy and early in pregnancy in ensuring successful pregnancy outcomes.12 Nationally representative data indicate that almost half of all US adults have what is considered poor diet quality.22 Studies focused on dietary intake among women before pregnancy or during early pregnancy show that women who are younger, non-Hispanic black, or of lower socioeconomic status have worse dietary intake than their reference groups.13,23,24 Previous studies tended to focus on small groups of women or examine a few nutrients or risk factors at a time. Studies of dietary intake during later pregnancy are more common than studies of dietary intake during early pregnancy, but few of these later-pregnancy studies are large scale and population based.25 If we are to improve maternal periconceptional nutrition, more extensive information on a wider variety of maternal characteristics and nutritional factors, across varied study populations, is needed.
Strengths of our study are that it included a large, diverse, population-based group of mothers who gave birth in 10 US states; therefore, results are generalizable to the US population. It is also unique in its focus on preconceptional diet and the range of nutritional factors and risk categories that were examined.
Limitations
Our study had several limitations. First, we did not adjust comparisons for any other factors; thus, the differences observed across groups may have been due to differences in other correlated factors (eg, we did not adjust the age results for race/ethnicity). Second, participation was 64%, which raises concerns about selection bias; however, a previous analysis that compared eligible control mothers who participated in the study with eligible control mothers who did not participate concluded that participants were generally representative of the underlying study populations (eg, in age, parity, smoking status).26 Third, data from the food frequency questionnaire were semiquantitative and were appropriate for ranking participants but not for determining actual intake; as such, we could not determine whether participants’ intakes were within actual recommended amounts. Fourth, the median time from delivery to interview was 7 to 8 months; as such, the 1 year before pregnancy could have started as far back as 29 months before the interview. We do not know the extent to which this amount of time affected the accuracy of recall.
Fifth, we conducted many comparisons yet used the conventional P < .05 as an indicator of significance; some of the observed significant differences in dietary intake may have arisen by chance, but we expect that this situation would not substantially affect the overall interpretation of our results. Sixth, some of the food groups we examined were broad (eg, we were not able to separate whole grains from grains); these broad groups parallel the construction of most dietary recommendations and were also constrained by what was asked in our instrument. Further consideration of the variability within these food groups would be useful for future studies; for example, examining factors such as the intake of low-fat dairy products, nonanimal protein, or dark green leafy vegetables. Seventh, we considered the possibility that dietary intakes may have changed during the 14-year study and that these changes may have contributed to the variability in the patterns of our results; however, this possibility seems unlikely. We compared average dietary intakes across 3 periods of data collection (1997-2001, 2002-2006, and 2007-2011) and found that dietary intakes were similar. We also reran the data for folate intake, restricting the analysis to the 9669 participants whose date of conception was January 1, 1999, or later. Folic acid fortification, mandatory as of January 1, 1998, would have been in place during the year before their pregnancy. Results for this subgroup of mothers were similar to the results for all study participants. The prevalence of risk factors may also have changed during the study period, but it is unlikely that their associations with nutritional intake would also have changed.
Conclusions
We identified subgroups of mothers who appeared to have worse nutritional intake than other groups of mothers before becoming pregnant. These subgroups included mothers who were aged <20; were nulliparous; were non-Hispanic black or had lower socioeconomic status, underweight BMI, or an unintended pregnancy; or who smoked. These subgroups are the very women already at elevated risk for adverse pregnancy outcomes. Poor nutritional status could exacerbate their risks. The information provided by this study may be useful for the development of preconceptional and periconceptional advice that helps ensure optimal nutrition among childbearing women.
Footnotes
Declaration of Conflicting Interests: The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported through cooperative agreements under PA 96 043, PA 02 081, FOA DD09-001, and RFA-DD-18-001 from the Centers for Disease Control and Prevention (CDC) to the Centers for Birth Defects Research and Prevention participating in the National Birth Defects Prevention Study. This project was partially supported by CDC U01DD001033 and U01/DD001226 and grant DK56350 from the University of North Carolina Department of Nutrition Clinical Research Center, Nutrition Epidemiology Core.
References
- 1. Karamanos B, Thanopoulou A, Anastasiou E, et al. Relation of the Mediterranean diet with the incidence of gestational diabetes. Eur J Clin Nutr. 2014;68(1):8–13. doi:10.1038/ejcn.2013.177 [DOI] [PubMed] [Google Scholar]
- 2. Tobias DK, Zhang C, Chavarro J, et al. Prepregnancy adherence to dietary patterns and lower risk of gestational diabetes mellitus. Am J Clin Nutr. 2012;96(2):289–295. doi:10.3945/ajcn.111.028266 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Grieger JA, Grzeskowiak LE, Clifton VL. Preconception dietary patterns in human pregnancies are associated with preterm delivery. J Nutr. 2014;144(7):1075–1080. doi:10.3945/jn.114.190686 [DOI] [PubMed] [Google Scholar]
- 4. Timmermans S, Steegers-Theunissen RP, Vujkovic M, et al. The Mediterranean diet and fetal size parameters: the Generation R Study. Br J Nutr. 2012;108(8):1399–1409. doi:10.1017/S000711451100691X [DOI] [PubMed] [Google Scholar]
- 5. Lowensohn RI, Stadler DD, Naze C. Current concepts of maternal nutrition. Obstet Gynecol Surv. 2016;71(7):413–426. doi:10.1097/OGX.0000000000000329 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Oken E, Ning Y, Rifas-Shiman SL, Rich-Edwards JW, Olsen SF, Gillman MW. Diet during pregnancy and risk of preeclampsia or gestational hypertension. Ann Epidemiol. 2007;17(9):663–668. doi:10.1016/j.annepidem.2007.03.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Qiu C, Coughlin KB, Frederick IO, Sorensen TK, Williams MA. Dietary fiber intake in early pregnancy and risk of subsequent preeclampsia. Am J Hypertens. 2008;21(8):903–909. doi:10.1038/ajh.2008.209 [DOI] [PubMed] [Google Scholar]
- 8. Brantsaeter AL, Haugen M, Samuelsen SO, et al. A dietary pattern characterized by high intake of vegetables, fruits, and vegetable oils is associated with reduced risk of preeclampsia in nulliparous pregnant Norwegian women. J Nutr. 2009;139(6):1162–1168. doi:10.3945/jn.109.104968 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Finer LB, Zolna MR. Declines in unintended pregnancy in the United States, 2008-2011. N Engl J Med. 2016;374(9):843–852. doi:10.1056/NEJMsa1506575 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Procter SB, Campbell CG. Position of the Academy of Nutrition and Dietetics: nutrition and lifestyle for a healthy pregnancy outcome. J Acad Nutr Diet. 2014;114(7):1099–1103. doi:10.1016/j.jand.2014.05.005 [DOI] [PubMed] [Google Scholar]
- 11. Shapira N. Prenatal nutrition: a critical window of opportunity for mother and child. Womens Health (Lond). 2008;4(6):639–656. doi:10.2217/17455057.4.6.639 [DOI] [PubMed] [Google Scholar]
- 12. Kind KL, Moore VM, Davies MJ. Diet around conception and during pregnancy—effects on fetal and neonatal outcomes. Reprod Biomed Online. 2006;12(5):532–541. [DOI] [PubMed] [Google Scholar]
- 13. Bodnar LM, Simhan HN, Parker CB, et al. Racial or ethnic and socioeconomic inequalities in adherence to national dietary guidance in a large cohort of US pregnant women. J Acad Nutr Diet. 2017;117(6):867–877. doi:10.1016/j.jand.2017.01.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Yoon PW, Rasmussen SA, Lynberg MC, et al. The National Birth Defects Prevention Study. Public Health Rep. 2001;116(suppl 2):32–40. doi:10.1093/phr/116.S1.32 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Willett WC, Sampson L, Stampfer MJ, et al. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol. 1985;122(1):51–65. [DOI] [PubMed] [Google Scholar]
- 16. US Department of Agriculture National Nutrient Database for Standard Reference, release 27. Nutrient Data Laboratory. http://www.ars.usda.gov/ba/bhnrc/ndl . Published 2014. Accessed November 20, 2018.
- 17. Bodnar LM, Siega-Riz AM. A diet quality index for pregnancy detects variation in diet and differences by sociodemographic factors. Public Health Nutr. 2002;5(6):801–809. doi:10.1079/PHN2002348 [DOI] [PubMed] [Google Scholar]
- 18. Haines PS, Siega-Riz AM, Popkin BM. The diet quality index revised: a measurement instrument for populations. J Am Diet Assoc. 1999;99(6):697–704. doi:10.1016/S0002-8223(99)00168-6 [DOI] [PubMed] [Google Scholar]
- 19. Carmichael SL, Yang W, Feldkamp ML, et al. Reduced risks of neural tube defects and orofacial clefts with higher diet quality. Arch Pediatr Adolesc Med. 2012;166(2):121–126. doi:10.1001/archpediatrics.2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. NHLBI Obesity Education Initiative Expert Panel on the Identification, Evaluation, and Treatment of Obesity in Adults. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults—the evidence report. Obes Res. 1998;6(suppl 2):51S–209S. [PubMed] [Google Scholar]
- 21. SAS Institute Inc. SAS [computer program]. Version 9.1. Cary, NC: SAS Institute Inc; 2004. [Google Scholar]
- 22. Rehm CD, Penalvo JL, Afshin A, Mozaffarian D. Dietary intake among US adults, 1999-2012. JAMA. 2016;315(23):2542–2553. doi:10.1001/jama.2016.7491 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Schaffer DM, Velie EM, Shaw GM, Todoroff KP. Energy and nutrient intakes and health practices of Latinas and white non-Latinas in the 3 months before pregnancy. J Am Diet Assoc. 1998;98(8):876–884. doi:10.1016/S0002-8223(98)00202-8 [DOI] [PubMed] [Google Scholar]
- 24. Rifas-Shiman SL, Rich-Edwards JW, Kleinman KP, Oken E, Gillman MW. Dietary quality during pregnancy varies by maternal characteristics in Project Viva: a US cohort. J Am Diet Assoc. 2009;109(6):1004–1011. doi:10.1016/j.jada.2009.03.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Blumfield ML, Hure AJ, Macdonald-Wicks L, Smith R, Collins CE. A systematic review and meta-analysis of micronutrient intakes during pregnancy in developed countries. Nutr Rev. 2013;71(2):118–132. doi:10.1111/nure.12003 [DOI] [PubMed] [Google Scholar]
- 26. Cogswell ME, Bitsko RH, Anderka M, et al. Control selection and participation in an ongoing, population-based, case-control study of birth defects: the National Birth Defects Prevention Study. Am J Epidemiol. 2009;170(8):975–985. doi:10.1093/aje/kwp226 [DOI] [PubMed] [Google Scholar]