Cleft lip with or without a cleft palate (CLP) is a serious birth defect, with varying birth prevalence rates among populations, gender, and geographic regions (International Clearinghouse for Birth Defects Monitoring Systems. Annual report 2003 with data for 2001. The International Centre for birth defects. 2003). These children have to undergo several treatments by various specialists, which have an enormous effect on their lives. Therefore, there is a strong imperative toward a better understanding of the cause of CLP, in which, in particular, maternal periconception exposures are of interest. Because harmful exposures can be avoided or treated, the preconception identification of risk factors will contribute to the primary prevention of CLP in next generations.
Evidence on the importance of maternal nutrition, lifestyle, and subtle genetic variants in the pathogenesis of CLP is increasing.1 In the first 12 weeks of gestation, the embryo is fully dependent on the supply by the mother for its nutrition. So far research on the association between maternal nutrition and CLP risk in the child has mainly been focused on associations between periconception intake of single nutrients, such as folate, vitamin B6, vitamin A, or zinc.2–4 Observational and intervention studies suggest a beneficial effect of periconception multivitamin supplementation on the occurrence of CLP.2,4,5 Of interest is that mothers with a compromised vitamin B-status either or in combination with a mild hyperhomocysteinemia, serving as a sensitive marker of B-vitamin status, demonstrate an increased risk of a child with CLP.2,6,7
The identification of dietary patterns has recently become of considerable interest and has been related to cardiovascular disease, type 2 diabetes, and cancer.8–10 So far, data on maternal dietary patterns and pregnancy outcome, in particularly birth defects, are lacking. Therefore, we investigated whether maternal dietary patterns derived from factor analysis are associated with 1) the risk of having a child with CLP and 2) the concentrations of folate, vitamin B12, vitamin B6, and homocysteine in maternal blood. Because it is not yet feasible to investigate these aims in a preconception cohort study, we have chosen as the best alternative conducting a case–control study at a fixed study moment approximately 14 months after the index-pregnancy. At that study moment, the nutritional status of the mother is rather comparable with the periconception period, because nutritional intake is stable during life except for periods of dieting, breastfeeding, and extreme growth.
MATERIALS AND METHODS
A case–control study was conducted that comprised of 225 case mothers of a child with a CLP or a cleft palate only and 217 control mothers of a nonmalformed child, as described in detail previously.4 All mothers were Dutch Europeans and studied at the standardized study moment of approximately 14 months after the delivery of the index child. Exclusion criteria were pregnancy, breastfeeding, folic acid–containing supplement use at the study moment, a different diet at the study moment than in the periconception period, and extreme nausea as hyperemesis or nausea starting after the first week of pregnancy. The institutional review boards of the participating hospitals approved the study protocol, and written informed consent was obtained from every participant.
We extracted the following data from the questionnaires: age, body weight, height, education, smoking, vitamin intake and other lifestyle factors. Body mass index (BMI) was calculated as the ratio weight (kg) per height (m2). Education was categorized into low (primary, lower vocational, intermediate secondary, or intermediate vocational), intermediate (higher secondary), and high education (higher vocational or university). Mothers were considered smokers when any smoking (cigarettes, cigars, or pipe) was reported. Data on folic acid only or multivitamin intake composed of content, dosage, and frequency.
The periconception period was defined as 3 months before until 3 months after conception of the index child. According to the Dutch periconception recommendation of folic acid supplementation, the periconception period for vitamin supplementation, however, was defined as the daily intake from 4 weeks before through 8 weeks after conception. Periconception vitamin intake and mild nausea or vomiting were included in the data set for analysis. Nausea was characterized by duration, period, and severity. Severe nausea or vomiting was defined as starting after the first week of pregnancy and resulting in a change or decrease in food intake.
The nutritional intake was assessed using a validated food frequency questionnaire.11,12 We standardly collected the nutritional intake at approximately 14 months after the index pregnancy. The rational was that nutritional habits in general are rather constant, with the exception of periods of dieting, pregnancy, and breastfeeding.13,14 The validity of the data is strengthened by the time at which the food frequency questionnaire was filled out, ie, 14 months after the index pregnancy, which is 24 months after the preconception period and in the same season as the preconception period. Moreover, most birth defects are diagnosed during the first postnatal year, which is important to reduce misclassification of the control group.
The food frequency questionnaire was mailed to the mothers and filled out at home. In the food frequency questionnaire the answers are given in frequency per day, week, month, year, or never. For several food items, additional questions were asked about the frequency of consumption and preparation methods, including the addition of condiments. The amounts consumed were estimated by using household measures or colored photographs of foods showing different portion sizes. The researcher verified the completeness and consistency of the food frequency questionnaire in a standardized way. Mean daily nutrient intake was estimated by multiplying the frequency of consumption of the food items by the portion size and the nutrient content per gram. Total energy intake, the intake of macronutrients, vitamins, minerals, and food groups were calculated by using the computerized version of the 1996 Dutch food composition table.15 The relative validity was assessed by comparing the data collected from this questionnaire with data drawn from 24-hour recalls, which has been repeated 12 times. The study demonstrated that the reproducibility and validity of food groups and nutrients was acceptable and comparable to other food frequency questionnaires12.
From 170 mothers a fasting venous blood sample was obtained to measure serum and RBC folate, serum vitamin B12, whole blood vitamin B6 as pyridoxal-'5-phosphate, and total plasma homocysteine.16–18 All laboratory analyses were performed blinded. Failure in blood sampling or laboratory testing resulted in approximately 5% missing values.
The procedure for calculating dietary patterns by using food consumption data from the food frequency questionnaire has been described, validated, and reproduced in detail elsewhere.19 To reduce the number of dimensions and to maintain the essence of the nutritional value of the foods, we reduced the number of food items (n=102) to 36 predefined food groups. The food groups were comparable to those reported by others.20,21 Food items that did not fit into any of the groups were considered as food groups on their own, eg, pizza, tea, and beer. After pooling of the case and control mothers, we performed principal factor analysis on all food groups. The Kaiser Meyer Olkin measure of sampling adequacy was used to exclude variables from the model due to multicollinearity.22 The number of factors to preserve was determined by factor interpretability and eigenvalues greater than 1. Bartlett's test of sphericity was used to test the equality of the eigenvalues.23 Correlation coefficients were analyzed by principal component analysis with subsequent varimax rotation to reveal the nutritional characteristics of each factor. The factor score for each pattern was calculated by adding up the intakes of the food groups weighted by the factor loadings. Each mother received a score for each factor solution to describe the similarity of her diet to the respective dietary pattern. The factor model was confirmed with reliability analysis. First, the set of 381 mothers was divided randomly into two groups. Factor analysis was performed on one group, and followed by the other group to confirm the model. The exploratory factor and reliability test were performed with SPSS 11.0.1 software (SPSS Inc., Chicago, IL).
Maternal age at the time of the study, BMI, and energy intake are presented as medians with 5th and 95th percentiles. Tertiles of the Western and Prudent dietary pattern were calculated based on factor scores. Differences between tertiles were evaluated by analysis of variance and tested for linearity. For each dietary pattern the differences in baseline characteristics were tested using the χ2 test. The tertiles of each dietary pattern of the total group of mothers and 85th percentile (n=58), 90th percentile (n=39), and 95th percentile (n=19) were used for risk estimation by odds ratios and 95% confidence intervals (95% CIs). Linear regression analysis was used to find associations between dietary patterns and nutrient intakes adjusted for energy intake.20 The nutrient residual method was used to adjust for total energy intake on micronutrient and macronutrient intakes.24 The distributions of the energy-adjusted nutrient intakes and maternal biomarker concentrations were positively skewed. Therefore, log transformation was performed, and the data are presented as geometric means with 5th and 95th percentiles. Differences between tertiles per dietary pattern were evaluated in an unconditional linear regression model. A multivariate model was used to analyze the effect of potential confounders, such as maternal education, smoking status, and alcohol consumption at the study moment.
After exclusion of 22 case mother and 39 control mothers, we evaluated the data of 203 case mothers and 178 control mothers. The CLP group comprised 171 mothers of a child with a CLP and 32 mothers of a child with a cleft palate only.
Five factors were selected from factor analysis. The first three factors represented typically Western dietary habits. Therefore, they were combined into one dietary pattern and called the Western dietary pattern. Factor 4 and factor 5 were combined and called the Prudent dietary pattern. The reliability test was positive and substantiates the validity of the data.
The first dietary pattern was called Western diet because of high intakes of organ meat, red meat, processed meat, pizza, legumes, potatoes, French fries, condiments, and mayonnaise, but low intakes of fruits. The Western diet explained 28.2 % of the total variance. The Prudent dietary pattern explained 7.1% of total variance and contained high intakes of fish, garlic, nuts, and vegetables and was characterized by a higher frequency of hot meals per day.
As the maternal Western dietary pattern score increased, the BMI generally increased as well (P<.01), maternal education decreased (P=.05), smoking increased (P<.001) and alcohol consumption decreased (P<.001) (Table 1). The Western dietary pattern is also characterized by higher intakes of saturated fats, monounsaturated and polyunsaturated fats, and cholesterol and lower intakes of total proteins, carbohydrates, fiber, β-carotene, ascorbic acid, thiamin, riboflavin, and pyridoxine.
Mothers in the highest tertile of the Western dietary pattern demonstrated a significantly increased CLP risk, odds ratio 1.9 (95% CI 1.2–3.1). After adjustment for maternal education, smoking and alcohol use at the study moment the increased risk for CLP remained significant. After adjusting the use of the Western dietary pattern for periconception folic acid or multivitamin intake, the significantly increased risk for CLP remained (Table 2). When considering the highest percentiles of the Western diet, a positive trend for CLP risk was determined ranging from an odds ratio of 2.0 (95% CI, 1.1–3.6) for the 85th percentile to 3.5 (95% CI, 1.1–10.7) for the 95th percentile with a P for trend of .024 (data not shown). On average, RBC folate, vitamin B12, and B6 concentrations decreased and total homocysteine levels increased with rising Western dietary pattern scores (Table 3).
The Prudent dietary pattern was not associated with increased CLP risk (Table 2). Mothers in the highest tertile of the Prudent diet were generally more highly educated (P<.05) and showed higher intakes of polyunsaturated fats, fiber, and β carotene. Average concentrations of vitamin B12 and folate increased with higher maternal Prudent dietary score (Table 3).
We stratified the dietary patterns for CLP and cleft palate–only offspring and no significant differences were revealed, P=.65 and .26, respectively. No significant differences were observed between the general characteristics of mothers of a CLP and a cleft palate only–child. The cleft palate only sample size was too small for further analysis.
Two major maternal dietary patterns have been identified in association with the risk of CLP offspring. These patterns have been labeled as the Western and Prudent dietary pattern, which is in line with others.8–10,20 In contrast to the Prudent diet users, mothers with the highest scores of the Western diet showed an approximately two-fold higher risk of having a child with a CLP, which was independent of periconception folic acid supplementation. The users of the Western diet were of a lower educational level and also demonstrated characteristics of an unhealthy lifestyle, such as a high BMI and smoking. These associations are in line with studies of others.8,10,20,21 The relations between dietary patterns and certain aspects of lifestyle are very interesting and may contribute to the future identification of specific risk profiles in the preconception counseling of mothers-to-be.
Maternal hyperhomocysteinemia is a risk factor for adverse outcome, such as a CLP.7 Studies on associations between dietary patterns and concentrations of B-vitamins and homocysteine are limited.20 Therefore, of interest are the associations between a high score for the Western pattern and high plasma total homocysteine, low RBC folate, low vitamin B12, and low pyridoxal-'5-phosphate concentrations that fit with a high intake of meat and low intakes of fruits and vegetables. The Prudent diet was positively associated with vitamin B12 and serum folate, very likely due to the high fish and vegetable consumption.
Our findings may suggest a significant effect of the maternal use of the Western dietary pattern on the risk of CLP offspring and herewith extend our previous findings from the multivariate analysis of food frequency questionnaire data.2,4,7 Those studies revealed adequate or slightly increased intakes of macronutrients and marginal intakes of several micronutrients in mothers of a CLP child compared with controls. The use of dietary patterns for identifying associations with lifestyle factors in combination with biomarkers and CLP risk is substantiated by these new findings and may improve the identification of a specific risk profile. Moreover, dietary patterns show several advantages above focusing on individual nutrients and foods, because synergistic effects between nutrients or foods may be more easily detected and the results seem to be more easily translated into practical advice to mothers-to-be.
A strength of the study is that we tested the data for validity and reproducibility by a reliability analysis. We divided the data set randomly into two halves, the principal component analysis was performed on each half, and the results were compared. The first factor loading matrix showed the same findings as the full data set. In the second factor loading matrix there was only one food group, ie, alcohol-free beer that did not match. This food group did not significantly affect the dietary patterns. Thus, the two analyses represent a study and a replication with valid results that support the validity of our findings.
The results reveal that the use of the Western dietary pattern during the preconception period may increase the risk of a child with a CLP. A limitation of our study is that we have not measured the nutritional intakes during the preconception period. This would require a very large preconception cohort study which is unfortunately not yet feasible. Therefore, we have chosen for the best alternative, namely, to study maternal exposures by conducting case–control studies at a fixed study moment at approximately 14 months after the index pregnancy to estimate the exposures in the preconception and periconception period.25–27 At that study moment the nutritional status of the mother is very likely comparable to the preconception and periconception period after excluding pregnant and breastfeeding women and those who have used a different diet at that moment compared with the preconception period. The rationale is that nutritional intake is rather stable during life except for periods of dieting, breastfeeding, and extreme growth. This is further supported by others.13,14
Our study shows that the concentrations of homocysteine, folate, vitamin B12, and B6 are significantly associated with the dietary patterns. Moreover, the validity is supported by a study on cardiovascular disease in the general Dutch population showing similar dietary patterns as in the mothers of our study. We have used other studies in nutritional epidemiology as guideline for our analysis to reduce possible bias.28 Nevertheless, the observed associations should be evaluated with respect to the results from food analysis. Bias in reported energy intake is associated with variable bias in estimated macronutrient intake. Moreover, environmental factors and genetic variations also affect embryogenesis and make it always difficult to detect association between CLP and dietary patterns. More studies, however, should be performed in the reproductive population to substantiate our data.
Current guidelines of preconception care emphasize that nutrition and certain lifestyle factors play an important role in early pregnancy. The identification in our study of the Western dietary pattern validated by some biomarkers and accompanied by unhealthy lifestyle factors may therefore be a first step in the profiling of preconception risks of mothers-to-be.
1. Krapels IP, Vermeij-Keers C, Muller M, de Klein A, Steegers-Theunissen RP. Nutrition and genes in the development of orofacial clefting. Nutr Rev 2006;64:280–8.
2. Krapels IP, van Rooij IA, Ocke MC, van Cleef BA, Kuijpers-Jagtman AM, Steegers-Theunissen RP. Maternal dietary B vitamin intake, other than folate, and the association with orofacial cleft in the offspring. Eur J Nutr 2004;43:7–14.
3. Munger RG, Sauberlich HE, Corcoran C, Nepomuceno B, Daack-Hirsch S, Solon FS. Maternal vitamin B-6 and folate status and risk of oral cleft birth defects in the Philippines. Birth Defects Res A Clin Mol Teratol 2004;70:464–71.
4. van Rooij IA, Ocke MC, Straatman H, Zielhuis GA, Merkus HM, Steegers-Theunissen RP. Periconceptional folate intake by supplement and food reduces the risk of nonsyndromic cleft lip with or without cleft palate. Prev Med 2004;39:689–94.
5. Shaw GM, Lammer EJ, Wasserman CR, O'Malley CD, Tolarova MM. Risks of orofacial clefts in children born to women using multivitamins containing folic acid periconceptionally. Lancet 1995;346:393–6.
6. van Rooij IA, Swinkels DW, Blom HJ, Merkus HM, Steegers-Theunissen RP. Vitamin and homocysteine status of mothers and infants and the risk of nonsyndromic orofacial clefts. Am J Obstet Gynecol 2003;189:1155–60.
7. Wong WY, Eskes TK, Kuijpers-Jagtman Spauwen PH, Steegers EA, Thomas CM, et al. Nonsyndromic orofacial clefts: association with maternal hyperhomocysteinemia. Teratology 1999;60:253–7.
8. Kerver JM, Yang EJ, Bianchi L, Song WO. Dietary patterns associated with risk factors for cardiovascular disease in healthy US adults. Am J Clin Nutr 2003;78:1103–10.
9. Rouillier P, Senesse P, Cottet V, Valleau A, Faivre J, Boutron-Ruault MC. Dietary patterns and the adenomacarcinoma sequence of colorectal cancer. Eur J Nutr 2005;44:311–8.
10. van Dam RM, Rimm EB, Willett WC, Stampfer MJ, Hu FB. Dietary patterns and risk for type 2 diabetes mellitus in U.S. men. Ann Intern Med 2002;136:201–9.
11. Ocke MC, Bueno-de-Mesquita HB, Goddijn HE, Jansen A, Pols MA, van Staveren WA, et al. The Dutch EPIC food frequency questionnaire. I. Description of the questionnaire, and relative validity and reproducibility for food groups. Int J Epidemiol 1997;26:S37–48.
12. Ocke MC, Bueno-de-Mesquita HB, Pols MA, Smit HA, van Staveren WA, Kromhout D. The Dutch EPIC food frequency questionnaire. II. Relative validity and reproducibility for nutrients. Int J Epidemiol 1997;26:S49–58.
13. Devine CM, Bove CF, Olson CM. Continuity and change in women's weight orientations and lifestyle practices through pregnancy and the postpartum period: the influence of life course trajectories and transitional events. Soc Sci Med 2000;50:567–82.
14. Leck I, Iles CA, Sharman IM, Toe T, Wadsworth GR. Proceedings: A study of the extent to which maternal nutrition after delivery reflects the situation in early pregnancy. Br J Prev Soc Med 1974;28:66.
15. NEVO Foundation. Dutch Food Composition Table 1996 [Dutch]. The Hague (Netherlands): Netherlands Nutrition Center; 1996.
16. Kuemmerle SC, Boltinghouse GL, Delby SM, Lane TL, Simondsen RP. Automated assay of vitamin B-12 by the Abbott IMx analyzer. Clin Chem 1992;38:2073–7.
17. Molloy AM, Scott JM. Microbiological assay for serum, plasma, and red cell folate using cryopreserved, microtiter plate method. Methods Enzymol 1997;281:43–53.
18. Schrijver J, Speek AJ, Schreurs WH. Semi-automated fluorometric determination of pyridoxal-5′-phosphate (vitamin B6) in whole blood by high-performance liquid chromatography (HPLC). Int J Vitam Nutr Res 1981;51:216–22.
19. Hu FB, Rimm E, Smith-Warner SA, Feskanich D, Stampfer MJ, Ascherio A, et al. Reproducibility and validity of dietary patterns assessed with a food-frequency questionnaire. Am J Clin Nutr 1999;69:243–9.
20. Fung TT, Rimm EB, Spiegelman D, Rifai N, Tofler GH, Willett WC, et al. Association between dietary patterns and plasma biomarkers of obesity and cardiovascular disease risk. Am J Clin Nutr 2001;73:61–7.
21. Lopez-Garcia E, Schulze MB, Fung TT, Meigs JB, Rifai N, Manson JE, et al. Major dietary patterns are related to plasma concentrations of markers of inflammation and endothelial dysfunction. Am J Clin Nutr 2004;80:1029–35.
22. Hutcheson GD, Sofroniou N. The multivariate social scientist: introductory statistics using generalized linear models. Thousand Oaks (CA): Sage Publications; 1999.
23. Pimentel RA. Morphometrics, the multivariate analysis of biological data. Dubuque (IA): Kendall-Hunt; 1979.
24. Willett WC, Howe GR, Kushi LH. Adjustment for total energy intake in epidemiologic studies. Am J Clin Nutr 1997;65:1220S–8S.
25. Groenen PM, van Rooij IA, Peer PG, Gooskens RH, Zielhuis GA, Steegers-Theunissen RP. Marginal maternal vitamin B12 status increases the risk of offspring with spina bifida. Am J Obstet Gynecol 2004;191:11–7.
26. Krapels IP, van Rooij IA, Ocke MC, West CE, van der Horst CM, Steegers-Theunissen RP. Maternal nutritional status and the risk for orofacial cleft offspring in humans. J Nutr 2004;134:3106–13.
27. Verkleij-Hagoort AC, Verlinde M, Ursem NT, Lindemans J, Helbing WA, Ottenkamp J, et al. Maternal hyperhomocysteinaemia is a risk factor for congenital heart disease. BJOG 2006;113:1412–8.
© 2007 The American College of Obstetricians and Gynecologists
28. Newby PK, Tucker KL. Empirically derived eating patterns using factor or cluster analysis: a review. Nutr Rev 2004;62:177–203.