- Email: [email protected]
Serum Folates in Gestating and Lactating Dairy Cows1
Serum Folates in Gestating and Lactating Dairy Cows 1 C. L. GIRARD and J. J. MA'n'E Agriculture Canada Research Station Lennoxville, Qu6bec J1M 1Z3, Canada G. F. TREMBLAY
Agriculture Canada Experimental Farm Normandin, Quebec G0W 2E0 Canada
A~.q'IRAOT In Experiment 1, 70 cows were distributed in five groups of 14 animals each. Each group represented one physiological stage: parturition, 2 mo postpartum, 3 mo of gestation, 6 mo of gestation, and drying off at approximately 2 mo before parturition. Plasma volume, concentration of serum folates, and total serum folates were measured at each stage. In Experiment 2, four doses of folic acid (40, 80, 160, and 320 mg) were administered by intramuscular injection to four groups of 5 cows in late gestation and four groups of 5 cows in early lactation. Serum folates in all cows and milk folates in lactating cows were determined before injection of folio acid and on d 1, 2, 4, 8, and 16 after injection. In Experiment 1, plasma volume did not differ between physiological stages, but total serum folates increased from parturition to reach a peak value 2 mo later; thereafter, serum folates decreased from 3 mo of gestation to parturition. In Experiment 2, during late gestation, serum folates increased after injection of folic acid and reached the highest concentration with the dose of 160 mg. However, during early lactation, injection of folic acid had no effect on concentrations of serum or milk folates. Therefore, total serum folates decreased by 40% from 2 mo postpartum (around mating) to parturition. Moreover, at the end of gestation, serum folates can be increased by an intramuscular injection of folic acid. This may be an indication of
ReceivedMarch 7, 1989. Accepted June 21, 1989. 1ContributionNumber264. 1989 J Dairy Sci 72:3240-3246
an increased need for folic acid during gestation of dairy cows. INTRODUCTION
The ruminant animal is generally considered independent of an exogenous supply of folates; its synthesis by rumen microorganisms makes their inclusion in the diet unnecessary (1, 26). According to Kon and Porter (18), folates content of rumen is not affected by the quantity of folates in the diet. However, differences in feeding practices seem to be responsible for the variation of concentrations of folates in milk; grazing herds have higher concentrations of folates in milk than herds fed dry hay (5). In previous studies, we observed that concentration of serum folates of calves aged 2 wk is half that of 4-too-old heifers (14). Moreover, a supplement of folic acid administered by intramuscular injections from the age of 2 to 18 wk improves growth performance (6). An increase of placental transfer of folates could have a beneficial effect on calf performance after birth. In humans, an increase in transfer of folates from the mother to the fetus takes place during the last weeks of pregnancy (7, 15). Fetal storage of folates increases in the second half of gestation in humans (19) and in rats (27). Furthermore, there is a relationship between maternal and newborn concentrations of folates in serum and whole blood in humans (7) and rats (34). If such is also the case in dairy cows, the folates status of the dam during gestation and lactation may be an indication of the wansfer of folates to the fetus. The objective of Experiment 1 was to observe the variations of serum folates during gestation and lactation in dairy cows. Fluctuations of serum folates not related to variations of plasma volume could indicate that, in some periods, requirements for folates are increased 3240
SERUM FOLATES 1N DAIRY COWS and not completely fulfilled. The objective of Experiment 2 was to verify if an intramuscular injection of folio acid, administered in late gestation or at the beginning of lactation, affected coneenla'ations of serum folates in dairy cows. MATERIALS AND METHOI~ Experiment 1
Animals. Seventy multiparous cows of the Agriculture Canada Research Station dairy herd at Lennoxville were randomly assigned to five groups of 14 animals each. Each group represented one physiological stage: parturition (0 to 1 d after parturition), 2 mo postpartum, 3 mo of gestation, 6 mo of gestation, and nonlactating cows approximately 2 mo before parturition. The cows were kept in a tie-stall barn under 12 h of light (0600 to 1800 h). They were milked twice daily at 0600 and 1630 h. A concentrate was fed three times daily (0800, 1100, and 1500 h) to meet NRC requirements (26) based on lactational performance of cows. Each cow received 4 kg hay/d divided into two meals (0800 and 1730 h). They also received oat silage for ad libitum intake twice daily (1000 and 1500 h). Body weight of cows was 628.0 (SE 9.6) kg. Blood Sampling. On sampling day, a disposable catheter (Angiocath, Deseret, Becton-Dickinson and Co., Sandy, UT) was inserted in the jugular vein. Twenty milliliters of blood w e r e drawn; 10 ml were kept for folates determination in serum and 10 ml were placed in heparinized tube for plasma volume determination. Ten milliliters of a solution containing 2% Evans blue dye (T1824; Sigma Chemicals, St. Louis, MO) diluted in a sterile saline (.9% NaCI) were injected in the catheter, the syringe and the catheter were carefully rinsed with blood to remove all dye. Ten minutes later, 10 ml of blood were drawn and placed in heparinized tube. Plasma Volume Determination. Plasma volume was measured in each cow with Evans blue dye as an indicator. Blood samples after injection of Evans blue dye solution were collected at fixed time (10 min) as reported for cows by Dalton and Fisher (4). Heparinized blood was centrifuged at 3296 × g for 20 rain and plasma collected. Dye concentrations were measured in fresh plasma on a
3241
LKB Ultrospec 2 spectrophotometer (Fisher, St. Foy, Quebec, PQ) at 615 nm. Optical density of plasma before injection of the dye served as a blank. Readings of optical density for each sample were reported on a standard curve to determine dye concentration in plasma. The standard curve was obtained in diluting known quantifies of Evans blue dye in a pool of cows' plasma. Plasma volume was calculated according to the Evans blue concentration in the plasma and the amount injected. Folates Determination. The blood was allowed to clot in the dark at room temperature for 2 h. The serum was separated by centrifugation at 1854 x g for 10 rain, transferred into polypropylene tubes, and stored at -20"C until assay. Serum folates were measured in duplicate by radioassay with a commercial kit used for human serum (Quantaphase Folate, Bio-Rad Laboratories (Canada) Ltd., Mississauga, ON.). To ascertain that the folates standard would apply to bovine serum, radioactive folic acid was added to sera collected from different cows and run as a nonspecific binding sample. The radioactivity counts for these bovine serum blanks were similar to those registered with the blanks routinely used in the procedure applied to human serum. Parallelism was good between 1 and 20 ng/ml; for values greater than 20 ng/ ml, the quantity of serum used in the assay was reduced. The interassay coefficient of variation was 5.05, and recovery tests with bovine serum using pteroylglutamic acid (ICN Canada Ltd., Montr6al, PQ) gave 107%. In view of these observations, it was concluded that this commercial kit could be accurately used for bovine serum analysis. Total serum folates were determined, for each cow, by multiplication of concentration of folates by plasma volume. Statistical Analysis. Values were analyzed as a completely random design (31) using the General Linear Model of SAS (32). The following model was used: Yi = I1, + Si + el, where Yi indicates the dependent variable, concentrations of serum folates, total serum folates, or plasma volume. The overall mean is bt and Si is the effect of the physiological stage. The variance analysis was followed by a Duncan's multiple range test (31) to test for significance of differences among means. Experiment 2
Animals. Twenty nonlactating cows approximately 6 wk before parturition and 20 lactating Journal of Dairy Science Voi. 72, No. 12, 1989
3242
GIRARD ET AL.
but nonpregnant cows 1 mo postpartum were used in Experiment 2. In each group, cows were randomly assigned to supplement of either 40, 80, 160, or 320 mg of folic acid. Housing conditions and feeding were the same as in Experiment 1. Body weight of cows was 625.6 (SE 10.9) kg. Folic Acid Administration. Each animal received on d 0 a single intramuscular injection of folic acid as a parenteral solution containing 40 mg pteroylglutamic acid/ml.
Blood Sampling and Folates Determination. Blood samples from the jugular vein were taken from each animal immediately before injection and 1, 2, 4, 8, and 16 d after the injection. Blood samples and injection of folic acid were made at 1300 h. The procedure for serum storage and folates determination was as described for Experiment 1.
Milk Sampling and Folates Determination. In lactating cows, milk samples were collected at the morning milking on the same days as blood samples. Milk was deproteinized by a boiling method adapted from Hoppner and Lampi (17). One milliliter of milk was diluted with 9 ml of phosphate buffer in a 15-ml polycarbonate tube for centrifugation, vortexed, and covered with aluminum foil. Fresh phosphate buffer was prepared every day (for 100 ml of phosphate buffer: .8709 g K2HPO,I, .20 g ascorbic acid, add distilled water, adjust pH at 7.8 with NaOH 1.0 N, and complete the volume at 100 mi with water). Tubes were placed in boiling water for 10 min, rapidly cooled on ice, vortexed, and centrifuged at 51,520 x g for 20 min. Supernatant was used for folates determination. Effect of chicken pancreas conjugase (transformation of polyglutamates to monoglutamates) on concentrations of folates was tested using two different methods (I 1, 17). There was no effect of conjugase on concentrations of milk folates, and subsequently, all assays were run without pretreatment with conjugase. These results seemed to confirm previous observations (29, 30) on the versatility of the radioassay technique for monoglutamates and polyglutamates. Parallelism was good between 1 and 20 ng/ml, and recovery tests were 104.2%. The interassay coefficient of variation was 4.76. Statistical Analysis. For each physiological stage, values were analyzed as a completely random design using the General Linear Model Journal of Dairy Science Vol. 72, No. 12, 1989
of SAS (32). The following model was used: Yij = ~t + Di + Tj + DTij + eij, where Yij indicates the dependent variable, serum, or milk folates. The overall mean is tx, Di is the dose effect, and Tj is the effect of time following the injection. According to Cox (3), the time response curves following the injection were divided in two parts: 0 to 1 d postinjection and 1 to 16 d postinjection. The effects of dose, time, and their interaction on concentrations of folates were split up into linear, quadratic, and other effects by orthogonal contrasts when appropriate (31). The analysis of repeated measurements were made according to Gill and Hafs (13). RESULTS Experiment 1
Plasma volume did not differ between physiological stages (P = .25). The effect of the physiological stage is more marked on total serum folates than on concentrations of serum folates (Table 1). Total serum folates increased from parturition to reach a peak value 2 mo later; thereafter, there was a decrease between 3 mo of gestation and parturition. Experiment 2
In nonlactating cows, mean concentration of serum folates was 13.56 (SE 1.01) ng/ml, on d 0. There was a dose effect of the injection of folic acid on concentration of serum folates (Figure 1; cubic effect of the dose, P = .0142); the highest concentration was after the injection of 160 mg folic acid. The time response was analyzed separately for both sides of the peak: d 0 to 1 postinjection (positive slope) and d 1 to 16 (negative slope). For all doses together, concentration of serum folates increased between d 0 and 1 (P = .0001) and thereafter decreased until d 16 (quadratic effect of the time, P = .0053). There was no interaction between dose and time (P = .27). In lactating cows, there was no effect of the dose of folic acid on concentrations of serum (Figure 1) and milk folates (P>.05). The effects of time and the interaction dose by time on serum and milk folates were not significant (P>.05). Mean concentration of serum folates was 23.31 (SE .64) ng/ml. Mean concentration
3243
SERUM FOLATES IN DAIRY COWS
TABLE 1. Plasmavolume,concentrationof sermn folates, and total serum folates in gestating and lactating dairy cows (n --- 14 cows per physiological stage). Physiological stage
Plasma volume
of folates
Total serum folates
Parturition1 2 mo Postpartum
(L) 30.03 33.93
(rig/nil) 16.88a 24.33b
(~g) 503.59a 836.63c
3 m o Gestation
30.98
21.91 b
6 mo Gestation 2 mo Prepartttm2 CV
30.88 30.58 15.48
17.64a 18.27a 23.06
685.24 b 541.83 ab
Concentration
566.55ab 31,9
a,b'eMeans within the same colunm with different superscripts differ (P<.05). 10 to 1 d after parturition. 2Nonlactatmg cows approximately 2 mo prepartum.
of milk folates was 59.02 (SE 1.51) ng/ml, varying from 30.55 to 128.18 ng/ml. DISCU.~.~ON
In Experiment 1, serum folates varied from 9.58 to 36.84 ng/ml and in Experiment 2, from 6.84 to 38.90 ng/ml; mean value was 18.84 ng/ ml for the two experiments. Arbeiter and Winding (2) report a mean concentration of 14.05 ng/ml in adult bovine animals with small individual variations. However, they do not specify these variations. Ford et al. (12) report concentrations varying from 16 to 39 ng/ml with a mean value of 26 ng/rnl. Lorin (20) reports a mean concentration of 14.5 ng/ml in multiparous cows varying from 9.65 to 21.46 ng/ml. In dairy heifers aged 4 mo, a concentration of 14.8 (SE .2) ng/ml was observed (14). These differences can be attributed to feeding practices or to variation in the synthesis of folates by rumen microflora. In Experimem 1, plasma volume was measured to be sure that a fall in concentration of serum folates was not principally due to plasma volume expansion like that reported in pregnant women (16). Plasma volume increased 2 mo after parturition but this increase was not significant. In pregnant heifers and cows at their first lactation, plasma volume increases during gestation and this increase is maintained during lactation (28). Mean plasma volume in Experiment 1 was 31.32 1 (SE 4.84) and represented 49.9 ml/kg body weight. This value was intermediate between the values reported previously in cows around parturition, 52.6 ml/kg body
weight (9), and in primiparous cows, 39.8 ml/ kg body weight (28). Concentration of serum folates was higher 2 mo after parturition and after 3 mo of gestation than in the other physiological stages studied. Arbeiter and Winding (2) report similar observations; gestating cows and cows in the postpartum period had lower concentrations of serum folates than cows non pregnant for at least 3 mo. However, the evolution of serum folates during gestation was not studied in theft experiment. In goats, the pattern seems to be different; concentration of serum folates is relatively stable during gestation and lactation except for a transient increase at parturition (12). However, in pigs and humans, where gestation and lactation are not concomitant, concentration of serum folates decrease during gestation, followed by a gradual rise until weaning (8, 23, 24 ). This pattern is similar to that observed in cows in the present experiment, where total serum folates reached a peak 2 mo after parturition in lactating, nonpregnant cows and thereafmr decreased to a minimum at parturition. In humans, lactation continues to drain mammal stores of folates and amplifies the risk of materhal deficiency when dietary folates intake is suboptimal during gestation and lactation (25). In dairy cows, the quantity and the quality of feed are usually higher after parturition and can help to restore maternal stores of folates. However, in spite of this, serum folates were already decreased after only 3 mo of gestation. Total serum folates decreased by 40% from around mating (2 mo after parturition) to parturition. In Experiment 2, nonlactating cows of all treatments exhibited an increase in concentraJournal of Dairy Science Vol. 72, No. 12, 1989
3244
GIRARD ET AL. 15
15
(01
14
14
E
E \
t2
u
c
o
9
I x
9
8
8
o
12 111
10 o
(b)
13
1]
7 6
I x
'll
m
o
7 6 4
3 0 k
1-
3 ¢ kl
04-11
-1
~ i1
-1
-2
-2 -3
I
0
2
4
6
8
10
12
14
16
18
2
I
1
4
I
I
6
I
I
8
I
I
10
I
I
12
I
I
14
I
I
16
]
18
I ~ kq~ ebT.CllOS
Figure 1. Variation in concentration of serum folatcs o f dairy cows relative to prcinjcction concentration (d x - d 0) after one intramuscular injection of 40 (~), 80 (+), 160 (¢), and 320 (A) m g of folic acid during a) late gestation (n = 20; SE .48) and b) early lactation (n = 20; SE .30).
tions of serum folates the day after an intramuscular injection of folic acid. Afterward, concentration of serum folates decreased gradually. The highest concentration of serum folates was observed after the injection of 160 rag. In lactating cows, an injection of folic acid had no effect on the concentration of serum folates. These effects are comparable to those observed in dairy heifers (14); injection of folic acid had a more marked effect in animals with a lower concentration of serum folates on d 0 than in those with a high concentration at the same time. The lack of response in lactating cows to intramuscular injection of folic acid may be explained by a rapid clearance of folic acid, a water-soluble vitamin. One day after injection, injected folic acid did not appear either in serum or in milk. Wagner (35) reported that high blood concentrations of folic acid promoted rapid renal excretion; because concentration of folates are lower in gestating cows, renal clearance may be reduced. The same phenomenon might appear for the highest dose Journal of Dah-y Science Vol. 72, No. 12, 1989
injected (320 mg folic acid) to gestating cows and explain why the peak observed on d 1 was lower than after a dose of 160 mg. The high concentration of serum folates reached after an injection of 320 mg folic acid might accelerate renal excretion of folates as compared with the other doses studied. The slower disappearance of folates after injection of folic acid in gestating cows than in lactating cows could also be partly explained by the presence in pregnant cows of folatebinding proteins. Folate-binding proteins were detected in cows' sera (12, 22). The presence of these proteins was also noted in human plasma and the binding of folates to those proteins in the maternal serum increases during pregnancy (21). Specific and nonspecific folate-binding proteins have been identified in the serum of pregnant women. Specific folate-binding proteins deliver folates to stores, predominantly to the liver, and represent a storage pool of folates. The nonspecific folate-binding proteins are responsible for the delivery of folates to the fetus, and probably also to sites of utilization.
3245
SERUM FOLATES IN DAIRY COWS
These folate-binding proteins can protect folates against renal excretion (10). Mean concentrations of total folates in raw c o w ' s milk treated with conjugase and dosed with Lactobacillus casei were 55 ng/ml (range 30 to 80 ng/ml) (12) and 77 ng/ml (range 40 to 118 ng/ml) (5). Milk folates measured in the present experiment by radioassay were similar to those reported in literature and were not affected by an injection of folic acid in early lactation. A supplement of folic acid had a small effect on concentration of milk folates in goats (12) but increased it in sows (24). In humans, a supplement of folic acid increased the concentration of milk folates in women receiving a low folates diet (25) but had no effect in well-nourished women (33). However, these reports demonstrate the effect of a chronic administration of a supplement and not of a ponctual supplement as used in the present experiment. In conclusion, serum folates in daily cows vary during gestation and lactation and under our feeding practices, serum folates decrease by 40% from 2 mo postpartum (around mating) to parturition. It seems that in dairy cows, generally considered independent of an endogenous supply of folates, the synthesis of folates by rumen microorganisms was not sufficient to prevent fluctuations of serum folates during gestation and lactation. Moreover, at the end of gestation, serum folates can be increased by an intramuscular injection of folic acid, but in cows during early lactation, these injections do not markedly affect serum and milk folates. Fetal storage of folates increases during the last part of the gestation in humans (7) and rats (27). If such is also the case in dairy cow, a supplement of folic acid administered during gestation could increase the quantity of folates available for the fetus. Therefore, further research is needed to determine the effect of maintaining a uniform concentration of serum folates during gestation and lactation on performances of dairy cows and their calves. Such research is already in progress in our laboratory. ACKNOWLEDGMENTS
The authors are grateful to C. Plante, M. Guillette, and M. Dumas for technical assistance, to the dairy barn staff at the Lennoxville
Research Station for taking care of the cows, and to L. C6t6 for preparation of the manuscript. This work was partially subsidized by Hoffmann-LaRoche, Basle, Switzerland and Mississauga, Ont., Can. REFERENCES 1 Agricultural Research Council. 1980. The nutrients requirements of ruminant livestock. Commonw. Agric. Bur., Slough. 2 Arbeiter, K., and W. Winding. 1973. Folatbestimmungen im Serum yon Rindem mit besonderem Bezug auf die Fruchtbarkeit. Wien. Tierarzl. Monatsschr. 60:323. 3 Cox, D. F. 1980. Design and analysis in nutritional and
physiological experimentation. J. Dairy Sci. 63:313. 4 Dalton, R. G., and E. W. Fisher. 1961. Plasma and blood volumes in Ayrshire cattle. Br. Vet. J. 117:115. 5 Dong, F. M., and S. M. Oace. 1975.Folate concentration and pattern in ovine milk. J. Agric. Food Chem. 23:534. 6 Dumouiin, P. G., C. L. Girard, J. J. Matte, and G. J. StLaurent. 1988. Effectsof intramuscular injection of folic acid on growth performances of young dairy heifers. J. Anita. Sci. 66(Suppl. 1):470. (Abstr.) 7 Ek, J. 1980.Pla.~raaand red cell folate values in newborn infants and their mothers in relation to gestational age. J. Peal. 97:288. 8 Ek, J. 1987. Folate and human milk. Pages 99-104 in
Human lactation. 3. The effects of human milk on the recipient infant. A. S. Goldman, S. A. Atkinson, and L. A. Hanson, ed. Plenum Press, New York, NY. 9Elley, D. S. 1979. Ire- and postpartum endocrine changes of the conccptus maternal unit in groups of Jersey cows selected for milk yield. M. S. Thesis, Univ. Florida, Gainesville. 10 Fernandes-Costa, F., and J. Metz. 1979. Role of serum folate binders in the delivery of fotate to tissues and to the fetus. Br. J. Hacmatol. 41:335. 11 Ford, J. E. 1967. The influence of the dissolved oxygen in milk on the stability of some vitamins towards heating and during subecqucnt exposure to sunlight. I. Dairy Sci. 34:239. 12 Ford, J. E., G. S. Knaggs, D. N. Salter, and K. J. Scott. 1972. Folate nutrition in the kid. Br. J. Nutr. 27:571. 13 Gill, J. L., and H. D. Hafs. 1971. Analysis of repeated measurements of animals. J. Anita. Sci. 33:331. 14 Girard, C. L., J. J. Matte, and G. L. Roy. 1989. Serum folates in young dairy heifers. Br. J. Nutr. 61:595. 15 Gross, S., B. Kamen, A. Fanaroff, and D. Caston. 1980. Folate compartments during gestational maturation. J. Fed. 96:842. 16 Hall, M. H., B.B.K. Pirani, and D. Campbell. 1976. The cause of the fall in serum folate in normal pregnancy. J. Obstet. Gynecol. Br. Commonw. 83:132. 17 Hoppner, K., and B. Lampi. 1981. Seasonal variation of folacin levels in market fluid milks. Can. Inst. Food Sci. Technol. 14:218. 18 Kon, S. K., and J.W.G. Porter. 1954. The intestinal synthesis of vitamins in the ruminant. Vitam. Horm. 12:
53. 19 Loria, A., A. Vaz-Pimo, P. Arroyo, C. Ramirez-Mateos, and L. Sanchez-Medal. 1977. Nu~tional anemia. VI. Journal of Dairy Science Vol, 72, No. 12, 1989
3246
G I R A R D ET AL.
Fetal hepatic storage of metabolites in the second half of pregnancy. L Pediaa'. 91:569. 20 Lorin, D. 1980. Folic acid in the blood. Concentration changes in connection with genital function. Pages 163-166 in Prec. 2rid Int. Syrup. Vet. Lab. D / a ~ s t i cians. Lucerne Switz. 21 Markkanen, T., P. Himanen, R.-L Pajula, S. Ruponen, and O. Castr6n. 1973. Binding of folic acid to serum proteins. The effect of pregnancy. Acta Haematol. 50:85. 22 Markkanen, T., R.-L. Pajula, P. Himanen, and S. Virtahen. 1974. Binding of folio acid to serum proteins. In some animal species. Int. J. Vitam. Nutr. Res. 44:34. 23 Matte, J. J., C. L. Girard, and G. J. Brisson. 1984. Serum folates during the reproductive cycle of sows. L Anita Sci. 59:158. 24 Matte, J. J., and C. L. Girard. 1989. Effects of intramuscular injections of folic acid during lactation on folates in serum and milk and performances of sows and piglets. J. Anita. Sci. 67:426. 25 Metz, J. 1970. Folate deficiency conditioned by lactation. Am. J. Clin. Nuu'. 23:843. 26 National Research Council. 1988. Nutrient requirements of dairy cattle. 0th ed. Natl. ?,cad. Sci., Washington, DC. 27 Potier de Courcy, G. et T. Tetroine. 1979. i~tude expOimeatale du d6veloppement foetal dans la d6ficience en acide folique induite par antivitamines
Journal of Dairy Science Vol. 72, No. 12, 1989
(acide X-methyl-folique, m6thotroxate). Ann. Biol. Anita Biochim. Biophys. 19:207. 28 Reynolds, M. 1953. Measurement of bovine plasma and blood volume during pregnancy and lactation. Am. J. Physiol. 175:118. 29 Rothenberg, S. P., M. daCosta, I. Lawson, and Z. Rosenberg. 1974. The determination of erythrocyte folate concentration using a two-phase ligand-binding radioassay. Blood 43:437. 30 Schreiber, C., and S. Wanman. 1974. Measurement of red cell folate levels b~ 3H-pteroylglutamic acid (3Hpteroylglutamic acid (H-Pte-Glu) radioassay. Br. J. Haematol. 27:551. 31Snedecor, G. W., and W. G. Cochran. 1971. M6thodes statistiques. 6th ed. Association Coord. Tech. Agric., Paris, Fr. 32 SAS Institute, Inc. 1985. User's guide. SAS Inst., Inc., Cary, NC. 33 Tamura, T., Y. Yoshimura, and T. Arakawa. 1980. Human milk folate and folate status in lactating mothers and their infants. An. J. Clin. Nutr. 33:193. 34 Thenen, S. W. 1979. Correlation between maternal and fetal folic acid status at day 21 of gestation in rats. Nutr. Rep. Int. 19:267. 35 Wagner, C. 1985. Folale-bincfing proteins. Nu~'. Rev. 43: 293.
Agriculture Canada Experimental Farm Normandin, Quebec G0W 2E0 Canada
A~.q'IRAOT In Experiment 1, 70 cows were distributed in five groups of 14 animals each. Each group represented one physiological stage: parturition, 2 mo postpartum, 3 mo of gestation, 6 mo of gestation, and drying off at approximately 2 mo before parturition. Plasma volume, concentration of serum folates, and total serum folates were measured at each stage. In Experiment 2, four doses of folic acid (40, 80, 160, and 320 mg) were administered by intramuscular injection to four groups of 5 cows in late gestation and four groups of 5 cows in early lactation. Serum folates in all cows and milk folates in lactating cows were determined before injection of folio acid and on d 1, 2, 4, 8, and 16 after injection. In Experiment 1, plasma volume did not differ between physiological stages, but total serum folates increased from parturition to reach a peak value 2 mo later; thereafter, serum folates decreased from 3 mo of gestation to parturition. In Experiment 2, during late gestation, serum folates increased after injection of folic acid and reached the highest concentration with the dose of 160 mg. However, during early lactation, injection of folic acid had no effect on concentrations of serum or milk folates. Therefore, total serum folates decreased by 40% from 2 mo postpartum (around mating) to parturition. Moreover, at the end of gestation, serum folates can be increased by an intramuscular injection of folic acid. This may be an indication of
ReceivedMarch 7, 1989. Accepted June 21, 1989. 1ContributionNumber264. 1989 J Dairy Sci 72:3240-3246
an increased need for folic acid during gestation of dairy cows. INTRODUCTION
The ruminant animal is generally considered independent of an exogenous supply of folates; its synthesis by rumen microorganisms makes their inclusion in the diet unnecessary (1, 26). According to Kon and Porter (18), folates content of rumen is not affected by the quantity of folates in the diet. However, differences in feeding practices seem to be responsible for the variation of concentrations of folates in milk; grazing herds have higher concentrations of folates in milk than herds fed dry hay (5). In previous studies, we observed that concentration of serum folates of calves aged 2 wk is half that of 4-too-old heifers (14). Moreover, a supplement of folic acid administered by intramuscular injections from the age of 2 to 18 wk improves growth performance (6). An increase of placental transfer of folates could have a beneficial effect on calf performance after birth. In humans, an increase in transfer of folates from the mother to the fetus takes place during the last weeks of pregnancy (7, 15). Fetal storage of folates increases in the second half of gestation in humans (19) and in rats (27). Furthermore, there is a relationship between maternal and newborn concentrations of folates in serum and whole blood in humans (7) and rats (34). If such is also the case in dairy cows, the folates status of the dam during gestation and lactation may be an indication of the wansfer of folates to the fetus. The objective of Experiment 1 was to observe the variations of serum folates during gestation and lactation in dairy cows. Fluctuations of serum folates not related to variations of plasma volume could indicate that, in some periods, requirements for folates are increased 3240
SERUM FOLATES 1N DAIRY COWS and not completely fulfilled. The objective of Experiment 2 was to verify if an intramuscular injection of folio acid, administered in late gestation or at the beginning of lactation, affected coneenla'ations of serum folates in dairy cows. MATERIALS AND METHOI~ Experiment 1
Animals. Seventy multiparous cows of the Agriculture Canada Research Station dairy herd at Lennoxville were randomly assigned to five groups of 14 animals each. Each group represented one physiological stage: parturition (0 to 1 d after parturition), 2 mo postpartum, 3 mo of gestation, 6 mo of gestation, and nonlactating cows approximately 2 mo before parturition. The cows were kept in a tie-stall barn under 12 h of light (0600 to 1800 h). They were milked twice daily at 0600 and 1630 h. A concentrate was fed three times daily (0800, 1100, and 1500 h) to meet NRC requirements (26) based on lactational performance of cows. Each cow received 4 kg hay/d divided into two meals (0800 and 1730 h). They also received oat silage for ad libitum intake twice daily (1000 and 1500 h). Body weight of cows was 628.0 (SE 9.6) kg. Blood Sampling. On sampling day, a disposable catheter (Angiocath, Deseret, Becton-Dickinson and Co., Sandy, UT) was inserted in the jugular vein. Twenty milliliters of blood w e r e drawn; 10 ml were kept for folates determination in serum and 10 ml were placed in heparinized tube for plasma volume determination. Ten milliliters of a solution containing 2% Evans blue dye (T1824; Sigma Chemicals, St. Louis, MO) diluted in a sterile saline (.9% NaCI) were injected in the catheter, the syringe and the catheter were carefully rinsed with blood to remove all dye. Ten minutes later, 10 ml of blood were drawn and placed in heparinized tube. Plasma Volume Determination. Plasma volume was measured in each cow with Evans blue dye as an indicator. Blood samples after injection of Evans blue dye solution were collected at fixed time (10 min) as reported for cows by Dalton and Fisher (4). Heparinized blood was centrifuged at 3296 × g for 20 rain and plasma collected. Dye concentrations were measured in fresh plasma on a
3241
LKB Ultrospec 2 spectrophotometer (Fisher, St. Foy, Quebec, PQ) at 615 nm. Optical density of plasma before injection of the dye served as a blank. Readings of optical density for each sample were reported on a standard curve to determine dye concentration in plasma. The standard curve was obtained in diluting known quantifies of Evans blue dye in a pool of cows' plasma. Plasma volume was calculated according to the Evans blue concentration in the plasma and the amount injected. Folates Determination. The blood was allowed to clot in the dark at room temperature for 2 h. The serum was separated by centrifugation at 1854 x g for 10 rain, transferred into polypropylene tubes, and stored at -20"C until assay. Serum folates were measured in duplicate by radioassay with a commercial kit used for human serum (Quantaphase Folate, Bio-Rad Laboratories (Canada) Ltd., Mississauga, ON.). To ascertain that the folates standard would apply to bovine serum, radioactive folic acid was added to sera collected from different cows and run as a nonspecific binding sample. The radioactivity counts for these bovine serum blanks were similar to those registered with the blanks routinely used in the procedure applied to human serum. Parallelism was good between 1 and 20 ng/ml; for values greater than 20 ng/ ml, the quantity of serum used in the assay was reduced. The interassay coefficient of variation was 5.05, and recovery tests with bovine serum using pteroylglutamic acid (ICN Canada Ltd., Montr6al, PQ) gave 107%. In view of these observations, it was concluded that this commercial kit could be accurately used for bovine serum analysis. Total serum folates were determined, for each cow, by multiplication of concentration of folates by plasma volume. Statistical Analysis. Values were analyzed as a completely random design (31) using the General Linear Model of SAS (32). The following model was used: Yi = I1, + Si + el, where Yi indicates the dependent variable, concentrations of serum folates, total serum folates, or plasma volume. The overall mean is bt and Si is the effect of the physiological stage. The variance analysis was followed by a Duncan's multiple range test (31) to test for significance of differences among means. Experiment 2
Animals. Twenty nonlactating cows approximately 6 wk before parturition and 20 lactating Journal of Dairy Science Voi. 72, No. 12, 1989
3242
GIRARD ET AL.
but nonpregnant cows 1 mo postpartum were used in Experiment 2. In each group, cows were randomly assigned to supplement of either 40, 80, 160, or 320 mg of folic acid. Housing conditions and feeding were the same as in Experiment 1. Body weight of cows was 625.6 (SE 10.9) kg. Folic Acid Administration. Each animal received on d 0 a single intramuscular injection of folic acid as a parenteral solution containing 40 mg pteroylglutamic acid/ml.
Blood Sampling and Folates Determination. Blood samples from the jugular vein were taken from each animal immediately before injection and 1, 2, 4, 8, and 16 d after the injection. Blood samples and injection of folic acid were made at 1300 h. The procedure for serum storage and folates determination was as described for Experiment 1.
Milk Sampling and Folates Determination. In lactating cows, milk samples were collected at the morning milking on the same days as blood samples. Milk was deproteinized by a boiling method adapted from Hoppner and Lampi (17). One milliliter of milk was diluted with 9 ml of phosphate buffer in a 15-ml polycarbonate tube for centrifugation, vortexed, and covered with aluminum foil. Fresh phosphate buffer was prepared every day (for 100 ml of phosphate buffer: .8709 g K2HPO,I, .20 g ascorbic acid, add distilled water, adjust pH at 7.8 with NaOH 1.0 N, and complete the volume at 100 mi with water). Tubes were placed in boiling water for 10 min, rapidly cooled on ice, vortexed, and centrifuged at 51,520 x g for 20 min. Supernatant was used for folates determination. Effect of chicken pancreas conjugase (transformation of polyglutamates to monoglutamates) on concentrations of folates was tested using two different methods (I 1, 17). There was no effect of conjugase on concentrations of milk folates, and subsequently, all assays were run without pretreatment with conjugase. These results seemed to confirm previous observations (29, 30) on the versatility of the radioassay technique for monoglutamates and polyglutamates. Parallelism was good between 1 and 20 ng/ml, and recovery tests were 104.2%. The interassay coefficient of variation was 4.76. Statistical Analysis. For each physiological stage, values were analyzed as a completely random design using the General Linear Model Journal of Dairy Science Vol. 72, No. 12, 1989
of SAS (32). The following model was used: Yij = ~t + Di + Tj + DTij + eij, where Yij indicates the dependent variable, serum, or milk folates. The overall mean is tx, Di is the dose effect, and Tj is the effect of time following the injection. According to Cox (3), the time response curves following the injection were divided in two parts: 0 to 1 d postinjection and 1 to 16 d postinjection. The effects of dose, time, and their interaction on concentrations of folates were split up into linear, quadratic, and other effects by orthogonal contrasts when appropriate (31). The analysis of repeated measurements were made according to Gill and Hafs (13). RESULTS Experiment 1
Plasma volume did not differ between physiological stages (P = .25). The effect of the physiological stage is more marked on total serum folates than on concentrations of serum folates (Table 1). Total serum folates increased from parturition to reach a peak value 2 mo later; thereafter, there was a decrease between 3 mo of gestation and parturition. Experiment 2
In nonlactating cows, mean concentration of serum folates was 13.56 (SE 1.01) ng/ml, on d 0. There was a dose effect of the injection of folic acid on concentration of serum folates (Figure 1; cubic effect of the dose, P = .0142); the highest concentration was after the injection of 160 mg folic acid. The time response was analyzed separately for both sides of the peak: d 0 to 1 postinjection (positive slope) and d 1 to 16 (negative slope). For all doses together, concentration of serum folates increased between d 0 and 1 (P = .0001) and thereafter decreased until d 16 (quadratic effect of the time, P = .0053). There was no interaction between dose and time (P = .27). In lactating cows, there was no effect of the dose of folic acid on concentrations of serum (Figure 1) and milk folates (P>.05). The effects of time and the interaction dose by time on serum and milk folates were not significant (P>.05). Mean concentration of serum folates was 23.31 (SE .64) ng/ml. Mean concentration
3243
SERUM FOLATES IN DAIRY COWS
TABLE 1. Plasmavolume,concentrationof sermn folates, and total serum folates in gestating and lactating dairy cows (n --- 14 cows per physiological stage). Physiological stage
Plasma volume
of folates
Total serum folates
Parturition1 2 mo Postpartum
(L) 30.03 33.93
(rig/nil) 16.88a 24.33b
(~g) 503.59a 836.63c
3 m o Gestation
30.98
21.91 b
6 mo Gestation 2 mo Prepartttm2 CV
30.88 30.58 15.48
17.64a 18.27a 23.06
685.24 b 541.83 ab
Concentration
566.55ab 31,9
a,b'eMeans within the same colunm with different superscripts differ (P<.05). 10 to 1 d after parturition. 2Nonlactatmg cows approximately 2 mo prepartum.
of milk folates was 59.02 (SE 1.51) ng/ml, varying from 30.55 to 128.18 ng/ml. DISCU.~.~ON
In Experiment 1, serum folates varied from 9.58 to 36.84 ng/ml and in Experiment 2, from 6.84 to 38.90 ng/ml; mean value was 18.84 ng/ ml for the two experiments. Arbeiter and Winding (2) report a mean concentration of 14.05 ng/ml in adult bovine animals with small individual variations. However, they do not specify these variations. Ford et al. (12) report concentrations varying from 16 to 39 ng/ml with a mean value of 26 ng/rnl. Lorin (20) reports a mean concentration of 14.5 ng/ml in multiparous cows varying from 9.65 to 21.46 ng/ml. In dairy heifers aged 4 mo, a concentration of 14.8 (SE .2) ng/ml was observed (14). These differences can be attributed to feeding practices or to variation in the synthesis of folates by rumen microflora. In Experimem 1, plasma volume was measured to be sure that a fall in concentration of serum folates was not principally due to plasma volume expansion like that reported in pregnant women (16). Plasma volume increased 2 mo after parturition but this increase was not significant. In pregnant heifers and cows at their first lactation, plasma volume increases during gestation and this increase is maintained during lactation (28). Mean plasma volume in Experiment 1 was 31.32 1 (SE 4.84) and represented 49.9 ml/kg body weight. This value was intermediate between the values reported previously in cows around parturition, 52.6 ml/kg body
weight (9), and in primiparous cows, 39.8 ml/ kg body weight (28). Concentration of serum folates was higher 2 mo after parturition and after 3 mo of gestation than in the other physiological stages studied. Arbeiter and Winding (2) report similar observations; gestating cows and cows in the postpartum period had lower concentrations of serum folates than cows non pregnant for at least 3 mo. However, the evolution of serum folates during gestation was not studied in theft experiment. In goats, the pattern seems to be different; concentration of serum folates is relatively stable during gestation and lactation except for a transient increase at parturition (12). However, in pigs and humans, where gestation and lactation are not concomitant, concentration of serum folates decrease during gestation, followed by a gradual rise until weaning (8, 23, 24 ). This pattern is similar to that observed in cows in the present experiment, where total serum folates reached a peak 2 mo after parturition in lactating, nonpregnant cows and thereafmr decreased to a minimum at parturition. In humans, lactation continues to drain mammal stores of folates and amplifies the risk of materhal deficiency when dietary folates intake is suboptimal during gestation and lactation (25). In dairy cows, the quantity and the quality of feed are usually higher after parturition and can help to restore maternal stores of folates. However, in spite of this, serum folates were already decreased after only 3 mo of gestation. Total serum folates decreased by 40% from around mating (2 mo after parturition) to parturition. In Experiment 2, nonlactating cows of all treatments exhibited an increase in concentraJournal of Dairy Science Vol. 72, No. 12, 1989
3244
GIRARD ET AL. 15
15
(01
14
14
E
E \
t2
u
c
o
9
I x
9
8
8
o
12 111
10 o
(b)
13
1]
7 6
I x
'll
m
o
7 6 4
3 0 k
1-
3 ¢ kl
04-11
-1
~ i1
-1
-2
-2 -3
I
0
2
4
6
8
10
12
14
16
18
2
I
1
4
I
I
6
I
I
8
I
I
10
I
I
12
I
I
14
I
I
16
]
18
I ~ kq~ ebT.CllOS
Figure 1. Variation in concentration of serum folatcs o f dairy cows relative to prcinjcction concentration (d x - d 0) after one intramuscular injection of 40 (~), 80 (+), 160 (¢), and 320 (A) m g of folic acid during a) late gestation (n = 20; SE .48) and b) early lactation (n = 20; SE .30).
tions of serum folates the day after an intramuscular injection of folic acid. Afterward, concentration of serum folates decreased gradually. The highest concentration of serum folates was observed after the injection of 160 rag. In lactating cows, an injection of folic acid had no effect on the concentration of serum folates. These effects are comparable to those observed in dairy heifers (14); injection of folic acid had a more marked effect in animals with a lower concentration of serum folates on d 0 than in those with a high concentration at the same time. The lack of response in lactating cows to intramuscular injection of folic acid may be explained by a rapid clearance of folic acid, a water-soluble vitamin. One day after injection, injected folic acid did not appear either in serum or in milk. Wagner (35) reported that high blood concentrations of folic acid promoted rapid renal excretion; because concentration of folates are lower in gestating cows, renal clearance may be reduced. The same phenomenon might appear for the highest dose Journal of Dah-y Science Vol. 72, No. 12, 1989
injected (320 mg folic acid) to gestating cows and explain why the peak observed on d 1 was lower than after a dose of 160 mg. The high concentration of serum folates reached after an injection of 320 mg folic acid might accelerate renal excretion of folates as compared with the other doses studied. The slower disappearance of folates after injection of folic acid in gestating cows than in lactating cows could also be partly explained by the presence in pregnant cows of folatebinding proteins. Folate-binding proteins were detected in cows' sera (12, 22). The presence of these proteins was also noted in human plasma and the binding of folates to those proteins in the maternal serum increases during pregnancy (21). Specific and nonspecific folate-binding proteins have been identified in the serum of pregnant women. Specific folate-binding proteins deliver folates to stores, predominantly to the liver, and represent a storage pool of folates. The nonspecific folate-binding proteins are responsible for the delivery of folates to the fetus, and probably also to sites of utilization.
3245
SERUM FOLATES IN DAIRY COWS
These folate-binding proteins can protect folates against renal excretion (10). Mean concentrations of total folates in raw c o w ' s milk treated with conjugase and dosed with Lactobacillus casei were 55 ng/ml (range 30 to 80 ng/ml) (12) and 77 ng/ml (range 40 to 118 ng/ml) (5). Milk folates measured in the present experiment by radioassay were similar to those reported in literature and were not affected by an injection of folic acid in early lactation. A supplement of folic acid had a small effect on concentration of milk folates in goats (12) but increased it in sows (24). In humans, a supplement of folic acid increased the concentration of milk folates in women receiving a low folates diet (25) but had no effect in well-nourished women (33). However, these reports demonstrate the effect of a chronic administration of a supplement and not of a ponctual supplement as used in the present experiment. In conclusion, serum folates in daily cows vary during gestation and lactation and under our feeding practices, serum folates decrease by 40% from 2 mo postpartum (around mating) to parturition. It seems that in dairy cows, generally considered independent of an endogenous supply of folates, the synthesis of folates by rumen microorganisms was not sufficient to prevent fluctuations of serum folates during gestation and lactation. Moreover, at the end of gestation, serum folates can be increased by an intramuscular injection of folic acid, but in cows during early lactation, these injections do not markedly affect serum and milk folates. Fetal storage of folates increases during the last part of the gestation in humans (7) and rats (27). If such is also the case in dairy cow, a supplement of folic acid administered during gestation could increase the quantity of folates available for the fetus. Therefore, further research is needed to determine the effect of maintaining a uniform concentration of serum folates during gestation and lactation on performances of dairy cows and their calves. Such research is already in progress in our laboratory. ACKNOWLEDGMENTS
The authors are grateful to C. Plante, M. Guillette, and M. Dumas for technical assistance, to the dairy barn staff at the Lennoxville
Research Station for taking care of the cows, and to L. C6t6 for preparation of the manuscript. This work was partially subsidized by Hoffmann-LaRoche, Basle, Switzerland and Mississauga, Ont., Can. REFERENCES 1 Agricultural Research Council. 1980. The nutrients requirements of ruminant livestock. Commonw. Agric. Bur., Slough. 2 Arbeiter, K., and W. Winding. 1973. Folatbestimmungen im Serum yon Rindem mit besonderem Bezug auf die Fruchtbarkeit. Wien. Tierarzl. Monatsschr. 60:323. 3 Cox, D. F. 1980. Design and analysis in nutritional and
physiological experimentation. J. Dairy Sci. 63:313. 4 Dalton, R. G., and E. W. Fisher. 1961. Plasma and blood volumes in Ayrshire cattle. Br. Vet. J. 117:115. 5 Dong, F. M., and S. M. Oace. 1975.Folate concentration and pattern in ovine milk. J. Agric. Food Chem. 23:534. 6 Dumouiin, P. G., C. L. Girard, J. J. Matte, and G. J. StLaurent. 1988. Effectsof intramuscular injection of folic acid on growth performances of young dairy heifers. J. Anita. Sci. 66(Suppl. 1):470. (Abstr.) 7 Ek, J. 1980.Pla.~raaand red cell folate values in newborn infants and their mothers in relation to gestational age. J. Peal. 97:288. 8 Ek, J. 1987. Folate and human milk. Pages 99-104 in
Human lactation. 3. The effects of human milk on the recipient infant. A. S. Goldman, S. A. Atkinson, and L. A. Hanson, ed. Plenum Press, New York, NY. 9Elley, D. S. 1979. Ire- and postpartum endocrine changes of the conccptus maternal unit in groups of Jersey cows selected for milk yield. M. S. Thesis, Univ. Florida, Gainesville. 10 Fernandes-Costa, F., and J. Metz. 1979. Role of serum folate binders in the delivery of fotate to tissues and to the fetus. Br. J. Hacmatol. 41:335. 11 Ford, J. E. 1967. The influence of the dissolved oxygen in milk on the stability of some vitamins towards heating and during subecqucnt exposure to sunlight. I. Dairy Sci. 34:239. 12 Ford, J. E., G. S. Knaggs, D. N. Salter, and K. J. Scott. 1972. Folate nutrition in the kid. Br. J. Nutr. 27:571. 13 Gill, J. L., and H. D. Hafs. 1971. Analysis of repeated measurements of animals. J. Anita. Sci. 33:331. 14 Girard, C. L., J. J. Matte, and G. L. Roy. 1989. Serum folates in young dairy heifers. Br. J. Nutr. 61:595. 15 Gross, S., B. Kamen, A. Fanaroff, and D. Caston. 1980. Folate compartments during gestational maturation. J. Fed. 96:842. 16 Hall, M. H., B.B.K. Pirani, and D. Campbell. 1976. The cause of the fall in serum folate in normal pregnancy. J. Obstet. Gynecol. Br. Commonw. 83:132. 17 Hoppner, K., and B. Lampi. 1981. Seasonal variation of folacin levels in market fluid milks. Can. Inst. Food Sci. Technol. 14:218. 18 Kon, S. K., and J.W.G. Porter. 1954. The intestinal synthesis of vitamins in the ruminant. Vitam. Horm. 12:
53. 19 Loria, A., A. Vaz-Pimo, P. Arroyo, C. Ramirez-Mateos, and L. Sanchez-Medal. 1977. Nu~tional anemia. VI. Journal of Dairy Science Vol, 72, No. 12, 1989
3246
G I R A R D ET AL.
Fetal hepatic storage of metabolites in the second half of pregnancy. L Pediaa'. 91:569. 20 Lorin, D. 1980. Folic acid in the blood. Concentration changes in connection with genital function. Pages 163-166 in Prec. 2rid Int. Syrup. Vet. Lab. D / a ~ s t i cians. Lucerne Switz. 21 Markkanen, T., P. Himanen, R.-L Pajula, S. Ruponen, and O. Castr6n. 1973. Binding of folic acid to serum proteins. The effect of pregnancy. Acta Haematol. 50:85. 22 Markkanen, T., R.-L. Pajula, P. Himanen, and S. Virtahen. 1974. Binding of folio acid to serum proteins. In some animal species. Int. J. Vitam. Nutr. Res. 44:34. 23 Matte, J. J., C. L. Girard, and G. J. Brisson. 1984. Serum folates during the reproductive cycle of sows. L Anita Sci. 59:158. 24 Matte, J. J., and C. L. Girard. 1989. Effects of intramuscular injections of folic acid during lactation on folates in serum and milk and performances of sows and piglets. J. Anita. Sci. 67:426. 25 Metz, J. 1970. Folate deficiency conditioned by lactation. Am. J. Clin. Nuu'. 23:843. 26 National Research Council. 1988. Nutrient requirements of dairy cattle. 0th ed. Natl. ?,cad. Sci., Washington, DC. 27 Potier de Courcy, G. et T. Tetroine. 1979. i~tude expOimeatale du d6veloppement foetal dans la d6ficience en acide folique induite par antivitamines
Journal of Dairy Science Vol. 72, No. 12, 1989
(acide X-methyl-folique, m6thotroxate). Ann. Biol. Anita Biochim. Biophys. 19:207. 28 Reynolds, M. 1953. Measurement of bovine plasma and blood volume during pregnancy and lactation. Am. J. Physiol. 175:118. 29 Rothenberg, S. P., M. daCosta, I. Lawson, and Z. Rosenberg. 1974. The determination of erythrocyte folate concentration using a two-phase ligand-binding radioassay. Blood 43:437. 30 Schreiber, C., and S. Wanman. 1974. Measurement of red cell folate levels b~ 3H-pteroylglutamic acid (3Hpteroylglutamic acid (H-Pte-Glu) radioassay. Br. J. Haematol. 27:551. 31Snedecor, G. W., and W. G. Cochran. 1971. M6thodes statistiques. 6th ed. Association Coord. Tech. Agric., Paris, Fr. 32 SAS Institute, Inc. 1985. User's guide. SAS Inst., Inc., Cary, NC. 33 Tamura, T., Y. Yoshimura, and T. Arakawa. 1980. Human milk folate and folate status in lactating mothers and their infants. An. J. Clin. Nutr. 33:193. 34 Thenen, S. W. 1979. Correlation between maternal and fetal folic acid status at day 21 of gestation in rats. Nutr. Rep. Int. 19:267. 35 Wagner, C. 1985. Folale-bincfing proteins. Nu~'. Rev. 43: 293.