Diet during pregnancy in a population of pregnant women in South West England

Transcription

1 European Journal of Clinical Nutrition (1998) 52, 246±250 ß 1998 Stockton Press. All rights reserved 0954±3007/98 $12.00 in a population of pregnant women in South West England I Rogers, P Emmett and the ALSPAC Study Team University of Bristol, Institute of Child Health, Unit of Paediatric and Perinatal Epidemiology, 24 Tyndall Avenue, Bristol BS8 1TQ, UK Objective: To describe the diet of a population of pregnant women. Subjects: Eleven thousand, nine hundred and twenty-three pregnant women resident in the south-west of England. Design: A self-completion unquanti ed food-frequency questionnaire was sent to the women at 32 weeks gestation. Estimated daily nutrient intakes were calculated from the answers to the questionnaire. Results: On the whole, nutrient intakes of the pregnant women in this survey compared very closely with the reported nutrient intakes for all women aged 16±64 in the last Dietary and Nutritional Survey of British Adults (DNSBA). The exceptions were sugar, calcium, folate and vitamin C where the estimated intakes were somewhat higher, and retinol, where the estimated intake was somewhat lower than the DNSBA. Mean and median estimated nutrient intakes were above the RNIs for the majority of nutrients investigated except for energy, iron, magnesium, potassium and folate. Levels of supplementary vitamin and mineral use were fairly low, the two most commonly taken supplements were ironðtaken by 22.5% of the pregnant women before 18 weeks and 43% at 32 weeks, and folate, taken by 9% and 18% of the women, respectively. Conclusions: These results suggest that the diets of pregnant women in this country are likely to contain adequate amounts of most nutrients, the most likely exceptions being iron, magnesium, potassium and folate. The relatively low intakes of folate and small proportion of women taking folate supplements is of concern, because of the association between inadequate amounts of folate in the diet and neural tube defects. Sponsorship: The nutritional aspects of the study have been supported by Northern and Yorkshire regionðnhs executive, Cow and Gate Ltd, the Meat & Livestock Commission and Coca-Cola UK. Descriptors: pregnancy; macronutrient intake; vitamin intake; mineral intake; folate Introduction Pregnancy and the periconceptional period is a time when the composition of the diet is of particular importance, as it may have a considerable in uence on birth outcome. Both excessive and inadequate intakes of certain vitamins (for example retinol and folate) may result in congenital defects (Bernhardt & Dorsey, 1974; Martinez-Frias & Salvador, 1990; Rothman et al, 1995; Bower & Stanley, 1989; MRC Vitamin Study Research Group, 1991; Czeizel & Dudas, 1992). Severe protein and energy shortages such as those experienced during the siege of Leningrad from 1941±1943 and the Dutch famine of the winter of 1944±1945 have been shown to result in a drop in birthweight, particularly when this occurs during the third trimester of pregnancy (Rush, 1982). A number of studies have suggested that even with more adequate energy and protein intakes, variations in the intakes of a large number of vitamins and minerals (Simmer et al, 1987; Doyle et al, 1990) and even specify fatty acids (Olsen et al, 1986, 1990) may have an effect on birthweight. Low birthweight is associated with an increased incidence of neonatal mortality (Bakketeig et al, 1984) and higher neonatal morbidity (Walther & Ramaekers, 1982), and possibly with an increased risk of developing a number of disorders in adult life (Barker & Fall, 1993; Barker et al, 1993a,b). Correspondence: Mrs P Emmett. Received 27 May 1997, revised 31 October 1997; accepted 9 November 1997 It is clearly desirable to have good up-to-date information on the average dietary intake of groups of pregnant women. However, there is a lack of recent large-scale dietary surveys in Britain which focus on this group. As part of the Avon Longitudinal Study of Pregnancy and Childhood we have collected information on diet in pregnancy from a group of over women. In this paper we present this data, and compare it to the nutritional recommendations for pregnant women contained in the 1991 Report of the Panel on Dietary Reference Values (Department of Health Report on Health and Social Subjects, 1991). Methods Study design The Avon Longitudinal Study of Pregnancy and Childhood (ALSPAC) is a geographically based cohort study investigating factors in uencing the health and development of infants and children. All pregnant women resident within that part of Avon, England which was within the South Western Regional Health Authority with an expected date of delivery between April 1991 and December 1992 inclusive were eligible. Women were recruited to the study mainly by direct approach from the booking midwife, but early recruitment was also obtained via advertisements in the media, and in doctors surgeries, maternity hospital booking clinics and chemist shops women enrolled in the study initially, representing 80±90% of the eligible population. Ethical approval of the study was obtained

2 from the ALSPAC ethics committee, and the local ethics committees of United Bristol, Southmead and Frenchay Healthcare Trusts. Information in the ALSPAC study is collected both from medical records and via a series of self-completion postal questionnaires sent to the mothers. Much of the information used in this paper is derived from a questionnaire sent out at 32 weeks gestation. This questionnaire was returned by women out of an estimated pregnancies reaching 32 weeks gestation. Diet was assessed by an unquanti ed food frequency questionnaire included within this, covering all the main foods consumed in Britain. The foods chosen for the questionnaire were based on those used by Yarnell et al, 1983 in South Wales and modi ed in the light of a more recent weighed dietary survey carried out by one of the authors on adults in Avon (Emmett et al, 1992). This included questions about the weekly frequency of consumption of 43 food groups and food items, the respondents were asked to tick one of the following optionsðnever or rarely, once in 2 weeks, 1±3 times a week, 4±7 times a week, more than once a day. More detailed questions were asked about daily consumption of a further eight basic foods (for example bread, coffee and tea, sugar). There were also questions about the types of certain foods used (for example cooking and spreading fats, soft drinks, milk and bread), and about the ways in which food was prepared and eaten, for example whether some or all of the fat was cut off meat, how often food was fried and how many of the slices of bread eaten in a day were spread with fat. In order to simplify the questionnaire and allow a more complex range of foods to be included no questions were asked about portion sizes, this is in contrast to the Yarnell questionnaire. Standard portion sizes were therefore used throughout. Alcoholic drinks were not included in this analysis, as the question about alcohol consumption was added one third of the way through the survey period, and so for many of the women it was not possible to estimate the weekly intake of each type of alcoholic drink. Food frequency questionnaires were chosen as the method of dietary assessment rstly because of the large number of subjects, and also because, while they do not provide such accurate quantitative information as weighed intakes, they give a reasonable estimate of the habitual diet (Bingham et al, 1994). The questionnaire used has not been validated in its present form. The effects of diet composition on pregnancy outcome could be modi ed by differences in the use of vitamin or mineral supplements. The questionnaire also asked about the use of various supplements in the previous three months. (A previous questionnaire had asked about supplement use in early pregnancy.) A copy of the food frequency questionnaire is available from the authors on request. Analysis of the food frequency questionnaire The food frequency questionnaire was used to calculate an approximate daily nutrient intake for each woman. This approximate daily intake was calculated by multiplying the weekly frequency of consumption of a food by the nutrient content (obtained from the 5th edition of McCance and Widdowsons `The Composition of Foods' and its supplements) (The Royal Society of Chemistry and MAFF, 1988, 1989, 1991a,b, 1992a,b, 1993, 1994) of a standard portion (Ministry of Agriculture Fisheries and Food, 1991) of that food, and summing this for all the foods consumed. The weekly frequencies of consumption assigned to each of the options ticked in the questionnaire were `never or rarely' ˆ 0, `once in 2 weeks' ˆ 0.5, `1±3 times a week' ˆ 2, 4±7 times a week' ˆ 5.5 and `more than once a day' ˆ 10. The question on bread consumption asked how many pieces of bread, rolls or chappatis were eaten on a normal dayðthe frequency options were less than 1, 1±2, 3±4 and 5 or more, which were assigned weekly consumption frequencies of 0, 10.5, 24.5 and 42 respectively. The level of milk consumption was calculated by summing the likely amount of milk drunk in tea and coffee, in breakfast cereal, in puddings and is milky drinks. The nutrient values obtained were then divided by seven to convert this to a daily intake. These approximate daily intakes were calculated for energy, protein, total fat, saturates, monounsaturates and polyunsaturates, total sugar, non-milk-extrinsic sugar, dietary bre (using Southgate analysis), nine vitamins and ve minerals. Results Population characteristics The age and anthropometric characteristics of the women for whom we obtained dietary information are shown in Table 1. Approximate daily intakes Dietary information was available for women. Table 2 shows the mean and median approximate daily intakes in comparison to the Reference Nutrient Intakes (RNIs) for pregnant women, and the mean nutrient intakes from food sources for all women aged 16±64 in the last Dietary and Nutritional Survey of British Adults (DNSBA) (Gregory et al, 1990). For the majority of nutrients examined the approximate daily intakes calculated from the food frequency questionnaire compared very closely with the values from the British Adults Survey. The exceptions were sugar, calcium, folate and vitamin C where the approximate daily intakes for the Avon women are somewhat higher than the DNSBA values, and retinol where the approximate daily intake was considerably lower. The approximate daily intakes for thiamine, ribo avin and pyridoxine were also a little higher than the DNSBA results. The mean and median approximate daily intakes were above the RNIs for the majority of nutrients, suggesting that the diet of the women in the ALSPAC survey is unlikely to be de cient in these nutrients. The exceptions were energy, iron, magnesium, potassium and folate. Use of vitamin and mineral supplements The proportion of women who had been taking the different vitamin and mineral supplements during early pregnancy, asked at 18 weeks of pregnancy, and in the three months before the administration of the dietary questionnaire at 32 weeks is shown in Table 3, in comparison with the level of supplement use recorded in the DNSBA. Only 17% of nonpregnant women took any form of supplement in the Table 1 Age and anthropometric characteristics of a population of pregnant women in South West England n Mean (sd) Minimum Maximum Age a (y) (5.0) Weight b (kg) (10.9) Height c (m) (0.07) BMI (kg/m 2 ) (3.8) a Age at delivery. b Self-reported pre-pregnancy weight. c Self-reported height. 247

3 248 Table 2 Approximate daily intakes from food sources only, as calculated from the food frequency questionnaire in comparison to the Reference Nutrient Intakes (RNIs) for pregnancy and results from the Dietary and Nutritional Survey of British Adulsts (DNSBA) for non-pregnant women aged 16±64 y Approximate daily intake Mean (sd) Median 5th %tile 95th %tile RNI a DNSBA b Energy (MJ) 7.70 (2.03) c 7.02 Protein (g) 66.3 (18.7) Fat (g) 70.4 (23.6) NA 73.5 Mono. fat (g) 24.8 (8.5) NA 22.1 Poly. fat (g) 12.5 (5.8) NA 11.0 Sat. fat (g) 29.1 (11.3) NA 31.1 Sugar (g) 105 (41) NA 86.0 NME sugar (g) 64 (37) NA NA Fibre (g) 19.4 (6.5) c 18.6 Calcium (mg) 953 (286) Total iron (mg) 10.4 (3.3) Haem iron (mg) 1.4 (0.8) NA NA Vegetable iron (mg) 9.0 (3.1) NA NA Total zinc (mg) 8.3 (2.4) d Meat zinc (mg) 2.1 (1.2) NA NA Vegetable zinc (mg) 6.2 (1.9) NA NA Magnesium (mg) 253 (76) d Potassium (mg) 2588 (661) d Retinol equivalents (mg) 855 (431) Carotene (mg) 2114 (1183) NA 2129 d Retinol (mg) 503 (365) NA 1058 Folate (mg) 250 (73) Niacin (mg) 15.8 (5.0) NA e NA e Riboflavin (mg) 1.73 (0.54) Thiamin (mg) 1.42 (0.42) Vitamin B6 (mg) 1.83 (0.53) Vitamin C (mg) 80.3 (34.7) Vitamin E (mg) 8.4 (4.3) NA 7.2 a The Reference Nutrient Intake shown in the table is the gure for women aged from 19±50, with the addition where appropriate of an increment for pregnancy. b The gures shown are the daily nutrient intakes from food sources only obtained by 7 d weighted record in the 1986±87 DNSBA. c The gure shown for energy and bre are the Estimated Average Requirements. d The gures shown for these nutrients are the intakes from all sources, as information on intake from food sources only was unavailable. e The RNI and the value from the DNSBA are for niacin equivalents (preformed niacin tryptophan=60), only values for preformed niacin are available from the ALSPAC survey. Table 3 Proportion of pregnant women (n ˆ 12104) taking dietary supplements during this pregnancy (asked at 18 weeks) and in the 3 months before 32 weeks gestation ALSPAC (%) ALSPAC (%) DNSBA a (%) Supplement before 18 weeks at 32 weeks non-pregnant Iron Zinc NA Calcium Folic acid NA `Vitamins' Other supplements NA a The gures given are for women of all ages in the Dietary and Nutritional Survey of British Adults. national survey, and for speci c supplements for which we had information the level of use was considerably higher among the pregnant women in Avon than in the national survey of non-pregnant women. Among pregnant women the use of iron and folate had increased from early to late pregnancy while the use of `vitamins' had decreased. Under-reporting of energy intake The number of women likely to be under-reporting was calculated by comparing the energy intake calculated from the food-frequency questionnaire with 120% of their Basal Metabolic Rate (BMR) calculated from their pre-pregnancy weight using Scho eld's equations (Department of Health Report on Health and Social Subjects, 1991), those women reporting an energy intake of less than 120% of BMR being considered to be under-reporters. By this criterion, 38% of the women in our sample under-reported energy intakeð this compares very closely with the DNSBA, where 39% of women reported an energy intake of less than 120% of BMR (Pryer et al, 1994). Discussion There are a number of aws in the unquanti ed food frequency method of assessing dietary intake. It may be dif cult for subjects to accurately judge the frequency with which they consume certain food items, particularly those which are consumed seasonally. There is no portion size information so standard portion sizes are assumed in the

4 calculation. This means that those women who eat large amounts of a small number of foods will have their intakes underestimated, while those who eat small amounts of a large variety of foods will have their intakes overestimated. Due to the absence of some nutrient values in the food tables it was necessary to choose foods according to whether or not they had nutrient values during the calculation of some approximate daily intakes, rather than choosing those foods which were felt to be most likely to be eaten. For example, for the majority of nutrients the foods used to calculate the nutrient content of a portion of green leafy vegetables were cabbage, Brussels sprouts and spring greens. However, to calculate the vitamin E content spinach was substituted for spring greens, as there is a missing value for the vitamin E content of spring greens in the food tables. In the weighed intake method the presence of missing values for some nutrients in some foods is rarely taken into consideration. The method of calculation of the approximate daily intakes assumes that the women eat a representative selection of the foods described in each question, whereas they may in fact only eat one of the foods in a group. In addition, some women may eat a signi cant number of foods which are not covered by the questionnaire, for example those women consuming ethnic diets. As a result there is a range of approximate daily intakes which is wider than a realistic range of nutrient intakes would be, although it is likely that the women would be ranked similarly by approximate daily intakes as by their actual nutrient intakes. The questionnaire cannot give a measure of diet at a speci c time point, which is especially important during pregnancy as any effect of diet on birth outcome may alter throughout the course of a pregnancy. However, there are also a number of advantages to the food frequency method of dietary assessment. For those nutrients where there is a large amount of day-to-day variability in the intakes (for example carotene or vitamin B 12 ) a food frequency questionnaire may provide a better measure of the usual intake than methods such as a few days of weighed records (Nelson et al, 1989). It places a far smaller burden on the subject than methods such as a seven-day weighed intake, and removes the temptation to alter the diet so as to make it easier to record. It is the only method which is really feasible for a survey of this size. Weighed intakes are often considered to be a `gold standard' of dietary assessment and several studies have obtained comparable results for nutrient intake when it has been assessed by both food frequency questionnaire and weighed intake. In a study in South West England (Emmett et al, 1992) assessing habitual intake of sugar, non-starch polysaccharide (NSP), starch, vitamin C and carotene using a food-frequency questionnaire and a four-day weighed intake, there was no signi cant difference between the average calculated nutrient intakes for NSP, intrinsic sugar and starch, and the differences in the calculated intakes of the other nutrients did not exceed 12%. A study of middle-aged women in Cambridge found that a food frequency questionnaire misclassi ed only 2±10% of individuals into opposite quartiles for nutrient intake when compared with the results of a 16-day weighed intake (Bingham et al, 1994). Food frequency questionnaires are also considerably less time-consuming for those administering the surveyðin the study by Emmett et al, 1992 administration and assessment of the food-frequency questionnaire took up approximately 30 min of the dieticians time, as opposed to 3 h for a 4-day weighed intake. The similarity of the results we have obtained using a food frequency questionnaire to those of the Dietary and Nutritional Survey of British Adults which used a seven-day weighed intake support their validity as an assessment of diet. In our calculations we have ignored the contribution of alcoholic drinks to nutrient intake. This is because the question on alcohol intake was not added to the questionnaire until seven months after the start of the survey therefore a substantial proportion of the women (29.8%) did not have opportunity to answer the question. The contribution of alcoholic drinks to nutrient intakes during pregnancy is likely to be very small on averageðamong those women (the nal 70% of the survey group) who did respond to a question on their levels of alcohol consumption at 32 weeks gestation more than two thirds reported consuming no alcohol at all. The approximate daily intakes for folate and vitamin C are higher than the mean intake values for women of all ages in the DNSBA, and the approximate intake for retinol is considerably lower. This may re ect dietary advice received by pregnant women, who are encouraged to include foods which are rich in folate and vitamin C in their diet, and to avoid foods which are concentrated sources of retinol such as liver (and also mould ripened cheeses, a relatively rich source of retinol). The approximate daily intakes for thiamine, ribo avin and pyridoxine are also a little higher than the DNSBA results. This may result from an increased consumption of forti ed breakfast cereals by pregnant womenðthese are frequently recommended as a good source of folate and to combat constipation. The approximate daily intakes for energy, iron, magnesium, potassium and folate fell below the RNIs for pregnant women. This suggests that these nutrients are the ones which are most likely to be inadequately supplied in the diets of the women in the ALSPAC survey. We have found a relatively low iron intake in pregnancy further research is needed to illuminate the effect of this on the health of the infant. A recent Japanese study showed that a group of 16 mothers with iron de ciency anaemia gave birth to infants with lower serum ferritin levels than those of non-iron de cient mothers (Hokama et al, 1996), however, although a signi cant proportion of the women were taking supplementary iron the study did not assess the diets of the women. Inadequate intakes of energy, particularly during the third trimester, have been associated with low birthweight (Rush, 1982). In this study intakes were only marginally lower than the recommendations and it is well-established that many people under-record their energy intakes using dietary assessment methods. We have estimated that 38% of the women in our survey probably under-recorded their energy intakes. Furthermore, there is some debate about how appropriate the current recommendations for energy intake during pregnancy are, some earlier studies (Durnin, 1987; Van Raaij et al, 1987) which estimated energy requirement during pregnancy found it to be considerably below the 1991 recommendation. These results highlight the possibility of inadequate folate intake in the periconceptional period which has been shown to increase the risk of delivering a baby with a neural tube defect (MRC Vitamin Study Research Group, 1991; Czeizel & Dudas, 1992). The intakes of many women at 32 weeks were below the RNIs, and only 18% were taking folate supplements. Even fewer of the women 249

5 250 (9%) were taking them in early pregnancy, so that in early pregnancy and the period shortly before conception, the time when women are recommended to increase their folate, intakes were likely to have been even lower. In 1996, after a health education campaign targeted at health professionals, only 30.6% of women were taking folic acid preconceptionally and 38.8% started supplements after conception, most of them too late for the folic acid to be effective (Wild et al. 1997). More than 30% of the ALSPAC pregnancies were unplanned, and women not planning a pregnancy were unlikely either to be taking folate supplements or to be deliberately including folaterich foods in their diets. The results of this survey seem to support the calls for forti cation of some staple foods in this country with folateðin the United States forti cation of our with folate is soon to be introduced (Schwarz & Johnston, 1996). As part of the ALSPAC study we have collected comprehensive and detailed information on other aspects of these women's lifestyles, and on the health and development of their children. These data thus provide an ideal opportunity for relating diet in pregnancy to aspects of pregnancy outcome, controlling for a wide range of confounding factors. AcknowledgementsÐWe are extremely grateful to all the mothers who took part in this study and to the midwives for their co-operation and help in recruitment. We would like to acknowledge the dedicated work of the ALSPAC study team, this includes interviewers, computer technicians, laboratory technicians, clerical workers, research scientists, volunteers and managers. The nutritional aspects of the study have been supported by Northern and Yorkshire regionðnhs executive, Cow and Gate Ltd, the Meat & Livestock Commission and Coca-Cola UK. The ALSPAC study is part of the WHO initiated European Longitudinal Study of Pregnancy and Childhood. References Bakketeig LS, Hoffman HJ & Titmuss Oakley AR. (1984): Perinatal mortality: In Perinatal Epidemiology, ed. Bracken MB, pp 99±151, New York, Oxford: Oxford University Press. Barker DJP & Fall CHD (1993): Fetal and infant origins of cardiovascular disease. Arch. Dis. Child. 68, 797±799. Barker DJP, Hales CN, Fall CHD, Osmond C, Phipps K & Clark PMS (1993a): Type 2 (non-insulin-dependent) diabetes mellitus, hypertension and hyperlipidaemia (syndrome X): relation to reduced fetal growth. Diabetologia 36, 62±67. Barker DJP, Osmond C, Simmonds SJ & Weild GA (1993b): The relation of small head circumference and thinness at birth to death from cardiovascular disease in adult life. Br. Med. J. 306, 422±426. Bernhardt IR & Dorsey DJ (1974): Hypervitaminosis A and congential renal anomalies in a human infant. Obstetrics and Gynecology 43, 750. Bingham SA, Gill C, Welch A, Day K, Cassidy A, Khaw KT, Sneyd MJ, Key TJA, Roe L & Day NE (1994): Comparison of dietary assessment methods in nutritional epidemiology: weighed records v. 24h recalls, food-frequency questionnaires and estimated-diet records. Br. J. Nutr. 72, 619±643. Bower C & Stanley FJ (1989): Dietary folate as a risk factor for neural tube defects: evidence from a case control study in Western Australia. Med. J. of Australia 150, 613±619. Czeizel AE & Dudas I (1992): Prevention of the rst occurrence of neuraltube defects be periconceptional vitamin supplementation. N. Engl. J. Med. 306, 1645±1648. Doyle W, Crawford MA, Wynn AHA & Wynn MA (1990): The association between maternal diet and birth dimensions. J. Nutr. Med. 1, 9±17. Department of Health Report on Health and Social Subjects (1991): Dietary Reference Values for Food, Energy and Nutrients for the United Kingdom. Durnin JVGA (1987): Energy requirements of pregnancy: an integration of the longitudinal data from the ve-country study. Lancet ii, 1131±1133. Emmett P, Symes C, Braddon F & Heaton K (1992): Validation of a new questionnaire for assessing habitual intakes of starch, non-starch polysaccharides, sugars and alcohol. J. Hum. Nutr. Dietet. 5, 245±254. Gregory J, Foster K, Tyler H & Wiseman M (1990): The Dietary and Nutritional Survey of British Adults. Hokama T, Takenaka S, Hirayama K, Yara A, Yoshida K, Itokazu K, Kinjho R & Yabu E (1996): Iron status of new borns born to iron de cient anaemic mothers. J Trop. Pediatr. 42, 75±77. Martinez-Frias, ML & Salvador J (1990): Aspects of prenatal exposure to high doses of vitamin A in Spain. Eur. J. Epidemiol. 6, 118±123. Ministry of Agriculture and Food (1991): Food Portion Sizes. London: HMSO. MRC Vitamin Study Research Group (1991): Prevention of neural tube defects: Results of the Medical Research Council Vitamin Study. Lancet 338, 131±137. Nelson M, Black AE, Morris JA & Cole TJ (1989): Between- and withinsubject variation in nutrient intake from infancy to old age: estimating the number of days required to rank dietary intakes with desired precision. Am. J. Clin. Nutr. 50, 155±167. Olsen SF, Hansen HS, Sorensen TIA, Jensen B, Secher NJ, Sommer S & Knudsen LB (1986): Intake of Marine Fat, Rich in Polyunsaturated Fatty Acids May Increase Birthweight by Prolonging Gestation. Lancet ii, 367±369. Olsen SF, Olsen J & Frische G (1990): Does Fish Consumption during Pregnancy Increase Fetal Growth? A Study of the Size of the Newborn, Placental Weight and Gestational Age in Relation to Fish Consumption during Pregnancy. Int. J. Epidemiol. 19, 971±977. Pryer JA, Vrijheid M, Nichols R, Kiggns M & Elliott P (1994): Who are the `low energy reporters' in the Dietary and Nutritional Survey of British Adults? Proc. Nutr. Soc. 53, 253A. Rothman KJ, Moore LL, Singer MR, Nguyen U-SOT, Mannino S & Milunsky A (1995): Teratogenecity of high vitamin A intake. New Engl. J. Med. 333, (21), 1369±1373. Rush D (1982): Effects of changes in protein and calorie intake during pregnancy on the growth of the human fetus. In: Effectiveness and Satisfaction in Antenatal Care, eds. M Enkin, I Chalmers, pp 92±113. London: Spastics International Medical Publications. Schwarz RH & Johnston RB (1996): Folic acid supplementationðwhen and how. Obstetrics and Gynecology 88, 886±887. Simmer K, Iles CA, Slavin B, Keeling PWN & Thompson RPH (1987): Maternal nutrition and intrauterine growth retardation. Hum. Nutr.: Clin. Nutr. 41C, 193±197. The Royal Society of Chemistry and MAFF (1988): Cereals and Cereal Products. The Royal Society of Chemistry and MAFF (1989): Milk Products and Eggs. The Royal Society of Chemistry and MAFF (1991a): The Composition of Foods. The Royal Society of Chemistry and MAFF (1991b): Vegetables, Herbs and Spices. The Royal Society of Chemistry and MAFF (1992a): Fruit and Nuts. The Royal Society of Chemistry and MAFF (1992b): Vegetable Dishes. The Royal Society of Chemistry and MAFF (1993): Fish and Fish Products. The Royal Society of Chemistry and MAFF (1994): Miscellaneous Foods. Van Raaij IMA, Vermaat-Miedema SH, Schonk CM, Peek MEM & Hautvast JGAJ (1987): Energy requirements of pregnancy in the Netherlands. Lancet ii, 953±955. Walther FJ & Ramaekers LHJ (1982): Neonatal morbidity of SGA infants in relation to their nutritional status at birth. Acta Paediatr. Scand 71, 437±440. Wild J, Sutcliffe M, Schorah CJ & Levene MI (1997): Prevention of neural-tube defects. Lancet 350, 30±31. Yarnell JWG, Fehily AM, Milbank JE, Sweetnam PM & Walker CL (1983): A short dietary questionnaire for use in an epidemiological survey: Comparison with weighed dietary records. Human Nutr.: Applied Nutr. 37A, 103±112.