Importance

Almost 50 years ago, interest in the relationship of vitamin B₆ to human pregnancy originated with the empirical use of pharmacologic doses of pyridoxine in the treatment of hyperemesis gravidarum—a condition of prolonged, severe nausea and vomiting during pregnancy (Willis et al., 1942). The lack of firm scientific evidence of the efficacy of this treatment is discussed later in this chapter. Evidence has accumulated that vitamin B₆ is required for protein, carbohydrate, and lipid metabolism as well as for erythrocyte, immune, and hormonal functions (see review by Leklem and Reynolds, 1988). Pyridoxal phosphate (PLP), the physiologically active form of the vitamin, is a coenzyme in over 100 known reactions involved primarily in amino acid metabolism. PLP-containing enzymes include aminotransferases, which are essential to the synthesis of nonessential amino acids, and decarboxylases, which are needed in the formation of histamine, serotonin, dopamine, and -aminobutyric acid. PLP is also a coenzyme in the formation of aminolevulinic acid, the first step in the synthesis of heme compounds. These vitamin B₆ -dependent reactions are of obvious importance to the normal course and outcome of pregnancy.

Estimated Requirements

Even though the vitamin B₆ intake and status of pregnant women have been widely studied, the requirements for this vitamin during pregnancy have not been clearly defined. It is known, however, that increased protein intake during pregnancy necessitates a modest increase in vitamin B₆ intake (Table 18-1), because of the major role of the vitamin in amino acid metabolism (NRC, 1989). Also, fetal uptake of vitamin B₆, especially in late pregnancy, increases the need for the vitamin. All forms of vitamin B₆, especially PLP, cross the placenta into fetal blood where concentrations are two to five times higher than those in maternal blood (Clearly et al., 1975; Contractor and Shane, 1970). Furthermore, the normal elevation of estrogen levels during pregnancy has been reported to increase tryptophan oxygenase activity (Rose, 1978), which in turn increases the need for vitamin B₆.

TABLE 18-1

Recommended Dietary Allowances of Water-Soluble Vitamins for Nonpregnant and Pregnant Women and the Rationale for Increased Allowances During Pregnancy.

A total body vitamin B₆ content of approximately 60 mg and a daily turnover rate of approximately 3% have been found in healthy nonpregnant women (Shane and Contractor, 1980). The vitamin B₆ content of blood was estimated to be less than 0.5 mg of pyridoxine equivalents, i.e., the concentration of individual B₆ vitamers calculated as pyridoxine. The amount of vitamin B₆ in maternal and fetal tissues gained has not been determined, but presumably represents only a small part of the estimated increased need for vitamin B₆ during pregnancy. The percentages of vitamin B₆ absorbed and metabolized to PLP as well as the oxidation and excretion of the vitamin appear to be the same during pregnancy as they are in the nonpregnant state (Contractor and Shane, 1970).

Decreases in both blood levels of vitamin B₆ and vitamin B₆ -dependent enzyme activity occur gradually during pregnancy. The most substantial decrease in plasma PLP levels is found between the fourth and eighth months of gestation, paralleling the period of most intensive growth of the fetus (Reinken and Dapunt, 1978). The fetus appears to lack the ability to phosphorylate pyridoxal and is dependent upon a maternal supply of PLP (Shane and Contractor, 1980). Thus, placental transport of PLP from mother to fetus is one mechanism that clearly leads to lower levels of PLP in maternal plasma, sometimes called the biochemical deficiency of vitamin B₆ of late pregnancy.

Criteria for Status Assessment

Overt clinical signs of vitamin B₆ deficiency (Table 18-2) are rare in the United States. In the absence of established markers, assessment procedures rely almost entirely on biochemical tests, including direct measurements B₆ of vitamers in blood or urine and indirect and functional tests to measure changes in PLP-dependent enzymes or activity coefficients (e.g., in vitro stimulation of enzyme activity by addition of PLP). Plasma PLP, which has been studied extensively, has been reported to be an indicator of vitamin B₆ body stores, whereas pyridoxic acid has been said to reflect intake (Leklem and Reynolds, 1988; Shane and Contractor, 1980; van den Berg, 1988). Concentrations of vitamin B₆ and B₆ vitamers in blood decrease with the normal increase in blood volume during mid- and late pregnancy.

An important consideration is the stage of pregnancy during which the tests are administered because of changes in hormonal balance throughout gestation. Such changes can affect enzyme turnover, enzyme-coenzyme binding properties, and redistribution of the B₆ vitamers in tissues. Knowledge about the effects of maternal homeostasis on the above-mentioned tests is limited (Shane and Contractor, 1980).

Studies of pregnant rats suggest that the pregnancy-induced changes in vitamin B₆ status indicators probably reflect a higher retentive capacity and temporary deposition of vitamin B₆ in tissues early in pregnancy as a result of hormone-induced changes (van den Berg and Bogaards, 1987). There is a need to quantify the influence of these secondary effects upon the biochemical indices of vitamin B ₆ status in pregnant women and then to set reference standards.

Studies have consistently shown that in comparison with nonpregnant controls, pregnant women have lower plasma levels of vitamin B₆ and PLP (Cleary et al., 1975; Contractor and Shane, 1970; Hamfelt and Tuvemo, 1972; Lumeng et al., 1976; Reinken and Dapunt, 1978; Roepke and Kirksey, 1979a; Schuster et al., 1984), decreased erythrocyte alanine aminotransferase activity, and higher activity coefficients (Lumeng et al., 1976; Schuster et al., 1981), especially during late pregnancy. Other changes include decreased levels of vitamin B₆ in leukocytes, erythrocytes, and urine and increased production of tryptophan or methionine metabolites following a large oral test dose (2 to 5 g) of the amino acid (Sauberlich, 1978). Abnormal results from a combination of two or more laboratory tests, e.g., decreased activity of vitamin B₆ -dependent enzymes coupled with high activity coefficients, are considered more indicative of vitamin B₆ inadequacy than is one abnormal measurement.

Usual Intakes

As shown in Chapter 13, Table 13-2, dietary intakes of vitamin B₆ by pregnant women in the United States have often been reported to be lower than the RDA (NRC, 1989). Using 3-day diet records, Roepke and Kirksey (1979a) calculated the mean daily vitamin B₆ intake of 97 middle-class U.S. women at 5 to 7 months of gestation to be 1.24 ± 0.55 (standard deviation [SD]) mg. Reynolds et al. (1984) analyzed the dietary intakes of 36 upper-middle-class U.S. women at 37 weeks of pregnancy and found their mean vitamin B₆ intake was 1.4 ± 0.42 (SD^* ) mg per day. The ratio of dietary vitamin B₆ to protein in these women was near the then recommended ratio for nonpregnant adults of 0.02 mg of vitamin B₆ to 1 g of protein. The current recommended ratio is 0.016 mg to 1 g (NRC, 1989). Among a group of 60 healthy pregnant Caucasian women, only three consumed 2.6 mg/day or more (Vir et al., 1980). The RDA at that time was 2.6 mg/day, compared with the current RDA of 2.2 mg/day (NRC, 1989). No significant relationship was observed between their vitamin B₆ status and the birth weights or anthropometric measurements of their neonates; however, the small sample size precluded definitive conclusions. In Florida, Schuster et al. (1981) reported that the mean daily vitamin B₆ intake of disadvantaged pregnant women (mostly of black origin) was 1.4 ± 1.0 (SD) mg, a level comparable to that reported for more economically advantaged women (Reynolds et al., 1984). The mean erythrocyte alanine aminotransferase activation coefficient among the Floridian women was 1.35 (compared with a normal value of ≤1.25 for nonpregnant women). Many values were considered by the researchers to be suggestive of vitamin B₆ inadequacy. In a subsequent study (Schuster et al., 1984) of 46 pregnant women from the same population, mean daily vitamin B₆ intake was estimated to be 1.5 mg (0.019 mg/g of protein).

Vitamin B₆ Status and the Course and Outcome of Pregnancy

Over the years, interest in the vitamin B₆ status of pregnant women has been stimulated by such findings as lower PLP concentrations in the umbilical cord blood of preeclamptic mothers compared with those in women with normal pregnancies (Brophy and Siiteri, 1975). However, Lu and colleagues (1981) failed to demonstrate any improvement in the course of toxemia following the administration of pyridoxine. Low levels of PLP in maternal plasma have been associated with low birth weight (Reinken and Dapunt, 1978), but this has not been uniformly confirmed (Vir et al., 1980).

Positive associations between vitamin B₆ status and the course and outcomes of pregnancy have been reported, but results of these studies are controversial, because no placebos were used and the subjects were not randomized or blinded. For example, early studies of pregnant women (Dorsey, 1949; Weinstein et al., 1944; Willis et al., 1942) in which pyridoxine doses of 5 to 100 mg/day were claimed to be effective in treating nausea and vomiting were not controlled; therefore, a placebo effect cannot be ruled out. In a study (Hesseltine, 1946) that included a placebo but that was not randomized or blinded, both pyridoxine and placebo were found to control nausea. The American Medical Association Council on Drugs (1979) has stated that there is no scientific evidence that vitamin B₆ is effective in the treatment of nausea. This viewpoint is supported by a recent review of the safety and efficacy of antiemetics in the treatment of nausea during pregnancy (Leathem, 1986). Associations of vitamin B₆ inadequacy with gestational diabetes (Spellacy et al., 1977) and with ''pregnancy depression''—described as pessimism, crying, tension without sleep, or appetite disorders (Pulkkinen et al., 1978)—have also been challenged on methodologic grounds.

The active transport of vitamin B₆ from maternal to fetal blood against a concentration gradient in the placenta lessens the effects of maternal vitamin B₆ inadequacy on the newborn, but it also could result in abnormally high levels in the fetus if pregnant women are given enough supplemental pyridoxine to increase their plasma PLP levels to those of nonpregnant women. Shane and Contractor (1975) postulated that this could adversely affect the synthesis of PLP-dependent enzymes by the fetus and might lead to a higher than normal vitamin B₆ requirement by the infant. However, this hypothesis has not been confirmed or refuted experimentally.

Three reports (Roepke and Kirksey, 1979a; Schuster et al., 1981, 1984) have related low vitamin B₆ intakes and low plasma levels as well as low PLP levels at delivery to unsatisfactory Apgar scores of newborns. These scores are based on heart rate, respiratory effort, muscle tone, reflex irritability, and color at 1 and 5 minutes after delivery (Apgar and James, 1962; Apgar et al., 1958), all of which can be influenced by many variables. Maternal pyridoxine supplementation was associated with improved Apgar scores taken at 1 minute; however, statistically significant improvements were not observed in 5-minute scores, which may be more indicative of long-term infant health problems (Schuster et al., 1984).

Since most pregnant women in the United States now consume multivitamin-mineral preparations containing vitamin B₆, it is usually not possible to conduct observational studies of relationships between dietary intake of vitamin B₆ and the course and outcome of pregnancy.

Toxicity

There are few data on the safety of pyridoxine supplementation during human pregnancy. Oral doses of pyridoxine greater than 500 mg/day for prolonged periods can result in the development of sensory neuropathy in nonpregnant adults (Cohen and Bendich, 1986) (Table 18-3). In the same review of the safety of pyridoxine, no toxic effects were reported for adults given 500 mg/day or less under medical supervision for periods ranging from 6 months to 6 years.

TABLE 18-3

Doses of Water-Soluble Vitamins Associated with Acute or Chronic Toxicity in Otherwise Healthy Pregnant and Nonpregnant Humans.

The suggestion that excessive vitamin B₆ intake during pregnancy produces a vitamin B₆ -dependency state in the newborn is based on one isolated case (Hunt et al., 1954). A woman treated with 50 mg of pyridoxine hydrochloride three or four times weekly for nausea during midpregnancy gave birth to an infant who had repeated convulsive seizures that responded to pyridoxine administration. The outcome was normal in an earlier pregnancy, during which the woman had not been given large doses of pyridoxine. Although pyridoxine-responsive convulsive disorders are occasionally observed in newborns, no reports have confirmed an association between them and maternal pyridoxine intake. Vitamin B₆ dependency appears to reflect an inborn error in metabolism rather than an acquired dependency state (Pitkin, 1982).

Recommendations for Supplementation

Most clinical trials of routine pyridoxine supplementation of pregnant women have failed to demonstrate any differences in pregnancy outcome, thereby casting doubt on the benefits of vitamin B₆ supplements. In a double-blind study, Schuster et al. (1984) found that a daily pyridoxine intake of 5.5 to 7.6 mg during pregnancy was required to avoid a decrease in plasma PLP levels at delivery. Without supplemental pyridoxine, mean levels decreased approximately 30% by 30 weeks of gestation and 25% at delivery over initial values. Lumeng et al. (1976) reported that pyridoxine intakes between 4 and 10 mg/day were needed to maintain plasma PLP at levels similar to those in the first trimester of pregnancy. Since physiologic changes during pregnancy may have accounted for the lower PLP levels, it is questionable whether pyridoxine supplementation should be used to produce levels similar to those in the nonpregnant state. Furthermore, the 4- to 10-mg/day doses of pyridoxine reported to maintain prepregnancy vitamin B₆ status during pregnancy exceed the amount obtainable from food.

For women at high risk for inadequate nutrient intake, e.g., substance abusers, pregnant adolescents, and women bearing multiple fetuses, the subcommittee recommends a daily multivitamin supplement containing 2 mg of vitamin B₆. This level is slightly less than the current RDA of 2.2 mg during pregnancy (NRC, 1989).

Long-term use (>30 months) of oral contraceptives containing high levels of estrogen (e.g., 100 g of mestranol or ethinyl estradiol) was associated with significantly lower maternal and umbilical cord serum vitamin B₆ levels than those in women who took no oral contraceptives, and evidence indicates that their vitamin B₆ reserves may be decreased in early pregnancy (Roepke and Kirksey, 1979b). Donald and Bossé (1979), Leklem (1986), and Leklem et al. (1975) concluded that oral contraceptive use, for short periods, does not significantly increase the need for vitamin B₆ . Concern has been expressed about women who routinely take oral contraceptives for several years to postpone their pregnancies (Miller, 1986). However, no data are available regarding the vitamin B₆ status of women taking the currently available oral contraceptives with low doses of estrogens (20 to 35 mg/day).

Importance

In India, more than 50 years ago, Wills (1931) successfully treated macrocytic anemia in pregnant women with yeast extract; the active substance was later identified as folate. The etiologic role of folate deficiency in megaloblastic anemia of pregnancy and the efficacy of folate therapy in the treatment of this disease are now well established. The fundamental roles of folate in cell replication and metabolism continue to be active areas of investigation.

Folates function in intermediary metabolism as coenzymes in the transfer of single carbon units (formyl, methyl, and formimino) from one compound to another. This step is vital to many metabolic processes, including the metabolism of several amino acids and the synthesis of purine and thymidylate—compounds essential to nucleic acid synthesis. In light of these fundamental roles of folate, a deficiency of this vitamin in the early weeks of pregnancy might be expected to impair cell growth and replication and to result in anomalies in the fetus and placenta, leading to subsequent spontaneous abortion, fetal malformation, or small-for-gestational-age infants (Hibbard, 1975). However, scientific evidence for these associations is inconclusive.

Folate Status and the Course and Outcomes of Pregnancy

Inconsistent results have been obtained in clinical studies to determine the association of mild to moderate folate deficiencies with spontaneous abortion, preterm delivery, fetal malformations, and low birth weight. This is due in part to imprecise definitions of folate status as well as methodologic weaknesses in some of the studies in which to placebos were used and subjects were not randomized or blinded. Some investigators (Hibbard, 1975; Iyengar and Rajalakshmi, 1975) have reported a high incidence of obstetric complications such as spontaneous abortions, toxemia, preterm and small-for-gestational-age infants, and antepartum hemorrhage in folate-deficient populations, whereas others (Giles, 1966; Scott and Usher, 1966) have failed to observe such relationships.

Adverse pregnancy outcomes have been linked with impaired folate status in disadvantaged populations in which folate deficiency and adverse birth outcomes are relatively common. In Johannesburg, South Africa, for example, an oral 500-µg/day supplement of folic acid was associated with a 50% reduction in small-for-gestational-age newborns among Bantu women consuming a low-folate diet; a similar effect was not observed among white women, who consumed more fruits and vegetables (Baumslag et al., 1970). In low-income, malnourished women in Hyderabad, India, oral supplementation with 500 µg of folic acid and 60 mg of elemental iron daily was associated with a 50% reduction in the number of low-birth weight infants (Iyengar and Rajalakshmi, 1975). A daily folate supplement of 500 µg was needed to maintain erythrocyte folate levels during pregnancy in Gambian women to ensure folate adequacy in the early stages of a subsequent pregnancy (Bates et al., 1986).

Unfortunately, several variables that might influence pregnancy outcomes were not always controlled in these studies. For example, subjects in the experimental groups were not uniformly controlled for age, weight, height, parity, previous pregnancy complications, socioeconomic status, prenatal care, or nutrient intake other than folate. There is no firm scientific evidence that the prophylactic use of folate lessens the complications or adverse outcomes of human pregnancy, with the exception of megaloblastic anemia.

Administration of folate antagonists such as aminopterin (4-amino folic acid) or methotrexate (methyl derivative of aminopterin) has consistently produced teratogenic effects in developing fetuses in both animals and humans. For example, when aminopterin was used as an abortifacient in humans, spontaneous abortions occurred in approximately 75% of the cases and the remaining fetuses were born grossly malformed (Goetsch, 1962; Thiersch, 1952, 1960), e.g., with fusion defects such as cleft lip and palate, hydrocephalus, and other major central nervous system deformities. The association of neural tube defects with folate and other nutrient deficiencies is discussed in Chapter 21.

Estimated Requirements for Pregnancy

Despite the crucial roles of folate in the synthesis of deoxyribonucleic acid (DNA) and in cell replication, the magnitude of the increased needs for folate during pregnancy has not been clearly defined. Dietary requirements rise with increased demands for the vitamin related to increased maternal erythropoiesis, uterine and mammary tissue growth, and placental and fetal growth. Requirements are further increased by greater urinary losses of the vitamin during pregnancy compared with those during nonpregnancy (Fleming, 1972; Landon and Hytten, 1971), but the percentage of folate absorption during pregnancy is unchanged (Iyengar and Babu, 1975).

The size of the folate body pool and equilibrium of the vitamin in relation to folate intake have not been assessed. Liver folate is a major portion of the body folate pool and the level of liver folate parallels the total body pool size (see review by Chanarin, 1979). Liver folate stores >3 µg/g (Hoppner and Lampi, 1980) and <1 µg/g (Gailani et al., 1970) have been suggested to reflect folate adequacy and deficiency, respectively. A dietary folate intake of 3 µg/kg of body weight has been reported to support adequate liver folate stores and to provide for a margin of safety in nonpregnant women (NRC, 1989; Reisenauer and Halsted, 1987).

To determine more precisely the magnitude of the increased folate need during pregnancy, a more complete understanding of cellular folate homeostasis and tissue folate requirements is required (Reisenauer and Halsted, 1987).

Criteria for Status Assessment

In advanced stages, folate deficiency is manifested as megaloblastic anemia, neutropenia, an increased number of hypersegmented polymorphs, and megaloblastic changes in bone marrow (Table 18-2). In earlier stages, these clinical signs may not be present, but the deficiency may be detected by biochemical indicators. The sequence of signs in the development of folate deficiency was observed in a healthy adult male placed on a folate-free diet (Herbert, 1962). In the first stage, serum folate dropped below normal levels. This was followed by an increased number of lobes on the nuclei of the polymorphonuclear leukocytes (hypersegmentation of neutrophils). After 4 months of folate deprivation, erythrocyte folate levels fell below normal and, subsequently, bone marrow became megaloblastic and anemia was evident.

TABLE 18-2

Biochemical Indices of Water-Soluble Vitamin Nutritional Status and Clinical Manifestations of Deficiency.

Measurements of folate levels in serum and erythrocytes are the most widely used biochemical indices of folate status. Low serum levels have been used to indicate depleted folate store (LSRO, 1984) but serve as a basis for treatment only when other signs have been observed. The decrease in serum folate levels during pregnancy has been partially attributed to blood volume expansion, increased urinary excretion of folate, and hormonal influences on folate metabolism. In some women, the continued depression of folate levels in serum and erythrocytes at 6 months postpartum suggested a chronic inadequacy of folate that failed to meet pregnancy needs and then was intensified postpartum (Bruinse et al., 1985). A study in Spain (Zamorano et al., 1985) showed that well-nourished, unsupplemented pregnant women did not have significant decreases in serum folate levels during their pregnancies.

Folate requirements increase rapidly during late pregnancy. This is reflected in decreased plasma folate levels, but not by the more slowly changing folate erythrocyte index (Chanarin, 1979). Nevertheless, researchers usually regard erythrocyte folate levels as the preferred indicator of folate status (Sauberlich, 1978). Erythrocyte folate levels are less sensitive than plasma indices to short-term changes in folate balance (Chanarin and Perry, 1977); a decrease in erythrocyte folate appears to reflect depletion of body folate stores. Neutrophil hypersegmentation (five or more nuclear lobes), ordinarily an early indicator of folate deficiency, is a poor indicator of folate status in pregnant women, since the number of lobes tends to decrease normally during pregnancy (Herbert et al., 1975).

Less commonly used assessment methods include folate functional tests, urinary excretion of formiminoglutamate, and the suppression of thymidine incorporation into DNA by deoxyuridine. Vitamin B₁₂ deficiency may, however, complicate these tests, since it interferes with normal folate metabolism. Measurement of the increase in reticulocytes in response to folate administration may be a useful indicator of folate status. Giles and Shuttleworth (1958) showed that folic acid supplements produced a peak increase in reticulocytes—from 0 up to 5 to 10% of circulating red blood cells—in most folate-deficient within 5 to 10 days after treatment.

Usual Intake

Folates are present in a variety of foods and occur in especially high levels in liver, fortified or whole grain breads and cereals, dried peas and beans, leafy vegetables, fruit (Subar et al, 1989), and yeast. In usual U.S. diets, most folate (approximately 75%) is found as polyglutamates (Butterworth et al., 1963). Human requirements can be met by a variety of chemical folate forms as long as the essential subunit structure of pteridine, p-aminobenzoic acid, and glutamic acid remains intact. If this structure is broken, biologic activity is lost. Heat, oxidation, and ultraviolet light can cleave the folate molecule, destroying its nutritional value. Thus, certain conditions of storage or cooking can reduce the folate content of foods.

Studies of the intestinal absorption of folate in humans show that monoglutamyl and polyglutamyl folate, the predominant forms in food, have similar bioavailabilities of about 50 to 70%; intestinal hydrolysis of polyglutamyl folate does not appear to limit its absorption (Chandler et al., 1986; Halsted, 1979; Halsted et al., 1986). Absorption of folate monoglutamate and folate polyglutamate was approximately 90% and 50 to 90%, respectively, in the absence of food intake but was lower when taken with various foods (Colman et al., 1975; Tamura and Stokstad, 1973). Food composition and intestinal absorption data suggest that the bioavailability of folate in typical U.S. diets is approximately one-half to two-thirds that of supplemental folic acid ingested separately from food (Sauberlich et al., 1987).

Estimates of dietary folate for population subgroups in the United States are limited. Furthermore, food composition data for folates are incomplete and uncertain. In the United States, the first large-scale dietary survey to include folate was the Continuing Survey of Food Intake by Individuals (CSFII), which began in 1985 (USDA, 1987). CSFII data obtained that year indicated that the mean folate intake by women between the ages of 19 and 34 (all income levels) was 217 µg daily. A special analysis of NHANES II data (Subar et al., 1989) found a mean folate intake of 206 µg daily by women in the same age group. These results are similar to the 227 µg/capita per day estimate of Anderson and Talbot (1981), which was based on the average per-capita use of principal U.S. foods. Ordinarily, per-capita estimates are high, since they have not accounted for food wastage and losses during cooking and storage.

In Boston, Huber et al. (1988) studied 566 pregnant women who were primarily white, middle class, and age 20 or older. Only 48 women in this group derived folate entirely from diet; mean folate intake of this small subgroup was 257 µg/day. In contrast, women who consumed folate supplements (91.5%) had a mean intake of 1,087 µg/day. Intake ranged as high as 6,759 µg/day, which is 16 times the 1989 RDA (NRC, 1989). In comparison to supplemented women, mean serum and erythrocyte folate levels in unsupplemented women were significantly lower, but no other evidence of folate inadequacy was reported.

Prevalence of Inadequacy

In the United States, folate deficiency has not been clearly identified as a general medical problem (Anderson and Talbot, 1981), nor has its prevalence among pregnant women been determined by biochemical or other indices. Colman et al. (1975) have nevertheless suggested that the added burden of pregnancy increases the potential risk and prevalence of folate deficiency. The estimated prevalence of compromised folate status depends, in part, upon the population subgroup studied and criteria used in making the diagnosis, as discussed below.

A high prevalence of folate inadequacy was reported among pregnant, low-income, predominantly black or Puerto Rican women living in New York City (Herbert et al., 1975) and in Florida (Bailey et al., 1980). In Florida, 48% of the study population was classified as folate deficient based on serum folate and 29% based on erythrocyte folate concentrations. In the New York sample, approximately 20% of the subjects were considered to be folate deficient based on serum folate and 16% on the basis of erythrocyte folate levels. No adjustments were made in these studies for the physiologic decline in these indices after midpregnancy. Furthermore, subjects with serum folate levels defined as deficient (i.e., <3 ng/ml) had no clinical manifestations of the deficiency.

Laboratory data on folate from the second National Health and Nutrition Examination Survey (NHANES II) (LSRO, 1984) show that within the U.S. population, the highest prevalence of low folate levels (serum level <3.0 ng/ml) and erythrocyte levels (<140 ng/ml) and, thus, the greatest risk of folate deficiency occurs among females (including a small number of pregnant women) aged 20 to 44. Of this population group, 15% had low serum folate, 13%, low erythrocyte folate, and 6%, both low serum and erythrocyte folate. The prevalence of low serum and erythrocyte folate was significantly greater among smokers than among nonsmokers and among nonusers of vitamin-mineral supplements than among users. Pregnancy, oral contraceptive use, and parity were also associated with low folate values among 22- to 44-year-old women.

Dosage Range and Toxicity

The safety of large doses of folic acid during pregnancy has not been systematically evaluated (Table 18-3). Low acute and chronic toxicity has been observed in nonpregnant adults. Folic acid is readily excreted in urine.

Large doses of folic acid may inhibit the absorption of other nutrients by competitive interaction (Ghishan et al., 1986; Simmer et al., 1987). They can also obscure the diagnosis of onset or relapse of pernicious anemia, which is extremely rare in women of childbearing age.

Ek (1980) has shown that plasma and erythrocyte folate levels in the fetus were two to four times higher than those in the mother. Since the effects of large maternal doses of folate on the developing fetus are not known, the subcommittee recommends that if folate supplements are used, they not exceed 300 µg/day as folic acid. This dose is higher than the total daily amount of folic acid needed by pregnant women (i.e., 200 µg/day), including those with poor folate stores, essentially no dietary folate, and multiple fetuses (Chanarin, 1985; Colman et al., 1975; NRC, 1989; Pritchard et al., 1969).

Recommendations for Supplementation

Pregnant women in the United States tend to consume less than the current RDA (NRC, 1989) of 400 µg of folate per day from food, have increased urinary losses of folate, and, especially if they are not supplemented, have a steady decrease in serum and erythrocyte folate levels. However, adequate folate for pregnancy can be obtained by regular consumption of fruits and vegetables in a well-selected diet (NRC, 1989). Thus, the subcommittee does not recommend routine folate supplementation during pregnancy but encourages daily use of fruits, vegetables, and whole grains and continued research to improve both the measurement of dietary folate and the assessment of requirements.

The subcommittee recommends modest supplementation for some segments of the U.S. population at risk of folate inadequacy, including some pregnant women who lack the knowledge or financial resources to purchase adequate food or who are abusers of alcohol, cigarettes, or drugs; or those who have malabsorption syndromes (LSRO, 1984). Pregnant adolescents and women bearing more than one fetus may also be at risk of folate deficiency. For these subpopulations, the subcommittee recommends folate supplements of 300 µg daily during pregnancy. This level has been recommended by several investigators (e.g., Chanarin, 1985; Colman et al., 1975; Hansen and Rybo, 1967; and Letsky, 1985).

Vitamin C

Thiamin, Riboflavin, and Niacin

Vitamin B₁₂

Vitamin B₁₂ is a group of cobalamins, i.e., cobalt-containing corrinoids with a tetrapyrrole structure resembling that of iron porphyrins. The predominant forms of the vitamin in plasma and tissues are methylcobalamin, adenosylcobalamin, and hydroxycobalamin. Cyanocobalamin is present in very small amounts in the body. Since it is the most stable form, it is used in vitamin supplements. Both cobalamin and folate function in the transport of single carbon atoms in reactions that are necessary for the synthesis of nucleic acids and the metabolism of certain amino acids. Thus, normal cell division and protein synthesis during pregnancy are dependent upon an adequacy of both vitamins.

Vitamin B₁₂ is supplied by animal protein foods, including meat, fish, eggs, and milk. The needs of pregnancy can be easily met by body stores or by diets that provide modest amounts of animal protein foods. Vegetarian diets that include eggs, milk, and cheese provide adequate vitamin B₁₂ for pregnancy needs (Immerman, 1981). Since a healthy fetus is estimated to contain about 50 µg, of vitamin B ₁₂, compared with maternal stores of approximately 3,000 µg, the drain on maternal stores for vitamin B₁₂ is usually slight (Immerman, 1981). The 1989 RDA of 2.2 µg/day during pregnancy (Table 18-1) is based on estimates of fetal needs of 0.1 to 0.2 µg/day and increased metabolism during pregnancy (NRC, 1989). An effective enterohepatic circulation recycles vitamin B₁₂ from bile and other intestinal secretions, accounting for its long biologic half-life.

Clinical deficiency of vitamin B₁₂ is usually secondary to abnormalities of gastrointestinal function. Deficiency caused by diet is very rare but is occasionally observed in adult vegans—complete vegetarians—who have followed an egg- and milk-free vegetarian diet for many years. If these individuals had previously consumed animal foods, their accumulated liver stores could protect them for several years (Immerman, 1981). In a few isolated cases, infants born to mothers who were complete vegetarians have manifested signs of vitamin B₁₂ deficiency during the first few months of life (Higginbottom et al., 1978; Sklar, 1986). In view of these findings, the subcommittee recommends a daily vitamin B₁₂ supplement of 2.0 µg for complete vegetarians.

Neither oral nor injectable cyanocobalamin has been found to be toxic to nonpregnant adults when administered in quantities several thousand times the daily requirement (LSRO, 1978), but the effects of excessive vitamin B₁₂ intake on the fetus have not been investigated.

Pantothenic Acid

Pantothenic acid is present in all living cells, mostly in the form of coenzyme A—an essential cofactor in the transfer of acetyl groups. A second active form of the vitamin is acyl carrier protein—a component of fatty acid synthetase complex. The vitamin is widely distributed in foods, especially in meats, whole-grain cereals, nuts, and legumes. Synthesis of pantothenic acid by intestinal bacteria possibly supplements the dietary intake of this vitamin. Spontaneous deficiency of pantothenic acid has not been observed in humans, except in cases of extreme malnutrition. Experimentally induced deficiency symptoms are intermittent diarrhea, insomnia, leg cramps, and paresthesias. Song et al. (1985) suggest that pregnant women need greater amounts of pantothenate than do nonpregnant women to maintain plasma levels and that such amounts are obtainable from food. Pantothenic acid toxicity in humans has not been reported. Occasional diarrhea is the only side effect reported to result from daily calcium pantothenate doses of 10 to 20 mg (Fox, 1984).

Biotin

Biotin is a sulfur-containing vitamin and a coenzyme for several important carboxylation reactions. Because it is synthesized by intestinal bacteria, spontaneous deficiency has not been observed in humans. A deficiency was produced experimentally in nonpregnant humans by feeding them large amounts of raw egg whites, which contain avidin—a biotin-binding protein. Symptoms of deficiency include seborrheic dermatitis, anorexia, muscle pain, and alopecia. Since biotin is widely distributed in food, needs are easily met by diet. Microflora synthesis also contributes to the biotin requirement. Blood levels of biotin fall progressively during pregnancy, but this has not been associated with adverse outcomes (Bonjour, 1984). No toxic effects of biotin were observed in nonpregnant humans following oral doses as high as 10 to 40 mg/day in the treatment of carboxylase deficiencies (Packman et al., 1981, 1985), but studies of toxicity during pregnancy have not been reported.