点击显示 收起
1 From the Department of Neonatology and Human's Nutrition Center, Hôpital Edouard Herriot, Lyon, France; the Laboratory of Biochemistry, Sainte Justine Hospital, Montréal; the Division of Child Health, University of Sheffield, Sheffield Children's Hospital, Sheffield, United Kingdom; and the Genetics Unit, Shriners Hospital for Children , Montréal.
2 Presented at the symposium Maternal Nutrition: New Developments and Implications, held in Paris, June 11–12, 1998. 3 Address reprint requests to BL Salle, Department of Neonatology, Hôpital Edouard Herriot, Place d'Arsonval 69437, Lyon Cedex 03, France. E-mail: salle{at}univ-lyon1.fr.
ABSTRACT
During pregnancy, maternal serum concentrations of 25-hydroxyvitamin D, the circulating form of vitamin D, correlate with dietary vitamin D intake. Maternal serum concentrations of 1,25-dihydroxyvitamin D, the hormonal circulating and active form of vitamin D, are elevated during pregnancy; 1,25-dihydroxyvitamin D is synthesized mainly by the decidual cells of the placenta and allows for increased calcium absorption. The fetus is entirely dependent on the mother for its supply of 25-hydroxyvitamin D, which is believed to cross the placenta. Hypocalcemia and increased parathyroid hormone secretion induce synthesis of 1,25-dihydroxyvitamin D after birth in both full-term and preterm neonates. Nevertheless, serum concentrations of 25-hydroxyvitamin D are a rate-limiting factor in the synthesis of 1,25-dihydroxyvitamin D. In vitamin D–replete infants, circulating 1,25-dihydroxyvitamin D concentrations are higher than those observed in older infants. In countries where dairy products are not routinely supplemented with vitamin D, maternal vitamin D supplementation during pregnancy is necessary. However, there is no indication for the use of pharmacologic doses of vitamin D or its metabolites in the perinatal period.
Key Words: 25-Hydroxyvitamin D • 1,25-dihydroxyvitamin D • mineral metabolism • parathyroid hormone • pregnancy • preterm infant • term infant
INTRODUCTION
The perinatal period, the subject of this review, includes the end of pregnancy and the first month of life in the newborn. Mineral homeostasis and vitamin D metabolism during this period have been the subject of several studies conducted over the past decade (1–6). The main questions addressed in this review are the following: 1) Does the fetus depend entirely on the mother for its supply of vitamin D metabolites? 2) Does the fetoplacental unit contribute to maternal and fetal needs for vitamin D metabolites, and if so, to what extent? 3) Does the vitamin D pool in the mother influence fetal mineral accretion? 4) Does the transient decrease in serum calcium during the first week of life in term and preterm infants represent adaptation to extrauterine life after removal of the placental source of minerals, inadequate vitamin D activation, or inappropriate sensitivity of target cells to calciotropic hormones? The effects of maternal vitamin D status on newborn infants and of postnatal vitamin D intake on metabolism and bone mineral accretion in both term and preterm infants are also discussed.
VITAMIN D METABOLISM
Cholecalciferol, or vitamin D3, is generally considered to be a prohormone synthesized in the skin after exposure to ultraviolet radiation; 7-dehydrocholesterol is converted to precholecalciferol and through thermal isomerization to cholecalciferol. Vitamin D2, or ergocalciferol, is produced by ultraviolet irradiation of the plant sterol ergosterol. Vitamins D2 and D3 are virtually equipotent in humans and can be included under the general name vitamin D. When present in the diet, fat-soluble vitamin D is efficiently absorbed with neutral lipids in the small intestine and transferred to the lymphatic system in chylomicrons. When formed in the skin, cholecalciferol is transported in blood, bound by an 2-globulin, to the liver to undergo further metabolism (7–9).
Vitamin D is cleared rapidly from the blood and lymphatics in the liver, where it undergoes a first hydroxylation at the carbon in position 25 to yield 25-hydroxyvitamin D, the main circulating form of vitamin D. Under normal circumstances, circulating concentrations of 25-hydroxyvitamin D can be regarded as a good index of nutritional vitamin D status. In humans, endogenous synthesis from sunshine exposure is an important source of vitamin D, but large seasonal variations in serum 25-hydroxyvitamin D concentrations have been observed in adults and infants (10). Bound to vitamin D binding protein, 25-hydroxyvitamin D is transported throughout the body and to the mitochondria of the renal cortex, where hydroxylation to a variety of metabolites takes place. The most important and biologically active metabolite, 1,25-dihydroxyvitamin D, plays a major role in maintaining appropriate blood calcium and phosphorus concentrations.
Decreased circulating concentrations of calcium increase serum parathyroid hormone (PTH) synthesis and secretion, which in turn stimulates the renal hydroxylation of 25-hydroxyvitamin D to 1,25-dihydroxyvitamin D. High concentrations of this metabolite may inhibit the hydroxylation of 25-hydroxyvitamin D and the secretion of PTH. 1,25-Dihydroxyvitamin D promotes the active intestinal absorption of calcium and phosphorus; the combined action of 1,25-dihydroxyvitamin D and PTH is to stimulate calcium mobilization from bones (7–9).
1,25-Dihydroxyvitamin D functions similarly to steroid hormones. The vitamin D receptor is present in the cytosol and has both ligand and DNA binding domains. Homodimerization or heterodimerization with the retinoic acid X receptor precedes the binding of the hormone receptor complex to specific vitamin D response elements located in the promoter regions of various genes that code for peptides such as collagen, osteocalcin, and PTH. Bone and intestine are traditionally regarded as the main target organs of 1,25-dihydroxyvitamin D action, but receptors have also been identified in tissues not directly involved in calcium metabolism, such as the endocrine glands, reproductive cells, and hematopoietic tissues. Furthermore, recent data suggest that some effects of 1,25-dihydroxyvitamin D may be transduced by cell surface receptors (11).
VITAMIN D METABOLISM DURING PREGNANCY
Maternal total serum calcium concentrations decline progressively throughout pregnancy and reach a nadir of 2–2.2 mmol/L by the second month. Because 50% of calcium is bound to serum albumin, hypoalbuminemia resulting from expansion of the extracellular volume accounts in part for this decrease; by contrast, serum ionized calcium concentrations undergo minimal changes (1, 2).
As alluded to above, serum 25-hydroxyvitamin D concentrations vary according to vitamin D intake and synthesis, season, and geographic location. Serum concentrations in pregnant women are either similar to or lower than those in nonpregnant women. Although it is difficult to assess whether liver hydroxylation of vitamin D is affected by pregnancy per se, modulation of different types of cytochrome P450 has been reported in this physiologic state. Hence, the possibility of altered liver hydroxylation should not be disregarded.
Circulating concentrations of 1,25-dihydroxyvitamin D are increased from the beginning of pregnancy (12–14). Hyperparathyroidism during pregnancy has been advocated as an explanation for this phenomenon. However, several studies showed that mean circulating immunoreactive PTH concentrations remained within the euparathyroid range (12, 15–17). Thus, a possible role for PTH remains controversial. In one study, serum immunoreactive PTH was measured in 14 healthy pregnant women during the last trimester by using immunometric assays specific for the intact hormone (amino acids 1–84), the midmolecule portion of PTH, and the carboxy terminal of PTH. As shown in Table 1, concentrations of all varieties of PTH remained within the normal range for healthy, nonpregnant women although 1,25-dihydroxyvitamin D concentrations increased dramatically. After birth, concentrations of the intact and carboxy terminal PTH did not change. In contrast, concentrations of the midmolecule PTH increased significantly 24 h after delivery (P < 0.025), but thereafter returned to previously recorded values.
View this table:
TABLE 1.. Mean serum calcium, phosphorus, 25-hydroxyvitamin D [25(OH)D], 1,25-dihydroxyvitamin D [1,25(OH)2D], osteocalcin, and immunoreactive parathyroid hormone (iPTH) concentrations in 14 healthy women during the last trimester of pregnancy and at delivery1
It was shown in rabbits that elevated maternal plasma concentrations of 1,25-dihydroxyvitamin D during gestation are in fact due to an increased production rate rather than decreased metabolic clearance (22). Whether, in this model, the increased 1,25-dihydroxyvitamin D synthesis is due to the mother's kidneys remains to be shown. In humans, a proportion of the circulating active metabolite was shown to be derived from maternal decidual cells (23–25). The kidney enzyme calcidiol 1-monooxygenase (25-hydroxycholecalciferol 1-hydroxylase) seems to escape the usual feedback mechanisms evoked earlier for the kidney, although it is most likely a product of the same gene as for the enzyme. Whether increased 1,25-dihydroxyvitamin D synthesis by the mother's kidneys indeed occurs remains to be shown; nevertheless, a proportion of the circulating active metabolite is derived from extrarenal sites such as maternal decidual cells.
Finally, it must be stressed that not more than 1% of vitamin D and its metabolites exist in plasma in the free state: the rest is bound to vitamin D binding protein. Serum concentrations of vitamin D binding protein increase during pregnancy. Nevertheless, toward the end of human pregnancy, there is an increase in the amount of free 1,25-dihydroxyvitamin D; at term, concentrations of free 1,25-dihydroxyvitamin D in plasma of mother and fetus are correlated (26).
Osteocalcin is a noncollagenous, bone-specific protein released by the osteoblast into the circulation in proportion to the rate of bone formation; the usefulness of osteocalcin measurement as a clinical marker of bone turnover is becoming widely accepted. Synthesis of osteocalcin by the osteoblast is stimulated by the action of 1,25-dihydroxyvitamin D and is influenced by PTH (27, 28).
As already reported by Cole et al (29) and Rodin et al (30), we observed that serum osteocalcin concentrations are low or undetectable from midpregnancy until delivery (Table 1). The cause of this decrease in osteocalcin concentrations during pregnancy remains elusive, although factors other than PTH or calcitriol are likely to be responsible. Farrugia et al (31) reported low concentrations of plasma osteocalcin during the late stage of pregnancy in ewes; after delivery, concentrations rose within 20 d postpartum. These authors also showed that maternal plasma osteocalcin concentrations were a direct reflection of the rate of entry of osteocalcin into the plasma pool because the plasma clearance rate did not differ significantly from that observed in nonpregnant controls. Thus, they excluded any significant contribution of the kidney or the placenta to the changes in circulating osteocalcin concentrations during normal ovine pregnancy. In addition, as shown in sheep and humans, there is no fetomaternal transfer of osteocalcin.
Using bone static histomorphometric measurements in a small number of women, Purdie et al (32) showed that bone turnover is diminished immediately postpartum. Although this observation might explain the decrease in serum osteocalcin concentration, the relation between osteoblast function, as indicated by osteocalcin production, and bone formation remains to be clarified in this particular setting. Seki et al (33) reported a correlation between serum osteocalcin and immunoreactive PTH concentrations at the end of pregnancy mainly as a result of a decline in serum PTH. These results are not corroborated by our findings because we found no correlation between osteocalcin and any of the variables reflecting mineral metabolism. In our experience, maternal serum osteocalcin concentrations increase sharply after delivery (Figure 1). Although no data are currently available in humans, we can hypothesize that low osteocalcin concentrations during pregnancy are due to placental trapping. Indirect evidence supporting this statement comes from the high osteocalcin content measured in the placenta (P Delmas, unpublished observations, 1992).
FIGURE 1. . Mean (±SEM) serum total 25-hydroxyvitamin D [25(OH)D], 1,25-dihydroxyvitamin D [1,25(OH)2D], and osteocalcin in 14 healthy pregnant women at delivery and 1 and 4 d after delivery. The women studied were the same subjects as in Table 1. A one-factor, repeated-measures ANOVA was performed, followed by a Scheffe F test.
The collective data suggest that under the influence of elevated serum concentrations of 1,25-dihydroxyvitamin D and appropriate calcium intake, calcium absorption is increased during pregnancy enough to more than offset the physiologic mechanism of calcium loss or transference to the fetus (ie, 20 g during the last trimester) (34). Indeed, Pitkin et al (1) showed in the early 1970s that malnourished pregnant women have negative calcium balance.
FETAL METABOLISM OF VITAMIN D
A major restriction in understanding fetal physiology is imposed by the relative inaccessibility of the human fetus. At present, biochemical analyses can only be performed at the time of birth. A strict time limitation exists, however, because mineral homeostasis changes rapidly in both mother and newborn in the first hours after delivery. New information may come from future studies based on the newly established cordocentesis techniques done during early pregnancy (10–20 wk of gestation).
Cord concentrations of the 3 major vitamin D metabolites are consistently lower than those measured in the mother's serum. Placental vein 25-hydroxyvitamin D and 24,25-dihydroxyvitamin D concentrations correlate significantly with those found in the maternal circulation, implying that these 2 secosteroids diffuse easily across the placental barrier and that the vitamin D pool of the fetus depends entirely on that of the mother (12, 35).
Fetomaternal relations of 1,25-dihydroxyvitamin D concentrations are more complex; in some studies, no correlation between fetal and maternal concentrations was observed whereas in other studies a highly significant correlation was found in both full-term and preterm infants (12). With pharmacologic administration of 1,25-dihydroxyvitamin D in the mother for treatment of hypoparathyroidism, transplacental transfer of 1,25-dihydroxyvitamin D may occur (36).
Congenital malformations of the kidney or viscera provide an opportunity to evaluate the relative contribution of the different compartments of the fetoplacental unit to vitamin D metabolite concentrations. Renal agenesis (Potter syndrome) is such a state, because the fetal renal tubular cell is the site of activity of calcidiol 1-monooxygenase. Studies show that most of the 1,25-dihydroxyvitamin D in fetal plasma is due to fetal kidney activity: cord blood concentrations of 1,25-dihydroxyvitamin D in fetal plasma from infants with Potter syndrome are one-third of those observed in healthy newborns (FH Glorieux, R Bouvier, JP Marie, et al, unpublished observations, 1984). Ross et al (37) also provided in vivo evidence for fetal synthesis of the hormone by showing that ovine fetal nephrectomy is associated with a substantial reduction in fetal plasma concentrations of 1,25-dihydroxyvitamin D.
FETOMATERNAL VITAMIN D RELATION
In countries in which vitamin D supplementation of milk is routine, vitamin D deficiency is unlikely to arise during pregnancy except in recent immigrants with chronic dietary insufficiencies of calcium, vitamin D, and other essential nutrients; in women avoiding dairy products for cultural or dietary reasons; or in women living where sunlight exposure is negligible (38). In malnourished populations with vitamin D deficiency, osteomalacia in the mother and abnormal skeletal metabolism in the fetus and infant have been reported (39). In other words, infants of severely malnourished mothers may be born with rickets and can suffer fractures in the neonatal period. Although much of the literature relates to specific populations, particularly the Bedouin population in Israel, reports from European and American centers indicate the need for continued vigilance, particularly in immigrant or refugee populations or in populations living in northern countries. In Berlin, 4 infants were born with craniotabes to osteomalacic immigrant mothers (39). The infants exhibited typical biochemical and radiologic changes that responded well to vitamin D therapy.
Observational studies suggest that radiographic bone density is reduced in both malnourished mothers and their infants (38). In addition, the results of several studies suggest that the bone mass of the newborn may be related to the vitamin D status of the mother. With use of dual-energy X-ray absorptiometry, it was shown that whole-body bone mineral content increases between 32–33 wk and 40–41 wk gestation (a 3.0-fold increase), suggesting that calcium accumulates in the fetus during the third trimester (40). A comparison of results from different countries shows that infant whole-body bone mineral content values are 20% lower in countries in which milk products are not supplemented with vitamin D (40, 41) than in countries where milk products are supplemented (42, 43). Recently, Namgung et al (44) reported that in Korea vitamin D status in winter is correlated with a marked reduction in total bone mineral content in newborns (Table 2).
View this table:
TABLE 2.. Influence of maternal vitamin D status on total-body bone mineral content (BMC) as measured by dual-energy X-ray absorptiometry in term infants1
Vitamin D nutrition in pregnancy was investigated by Brooke et al (45) in 115 Asian (Pakistani, Hindu Indian, and East African Asian) women living in London and in 50 of their newborn infants. The mean maternal serum 25-hydroxyvitamin D concentration at the beginning of the last trimester was 20.2 nmol/L and decreased to 16.0 nmol/L after delivery. Postpartum, 36% of the women and 32% of the infants had undetectable 25-hydroxyvitamin D concentrations (<3 nmol/L). Alkaline phosphatase bone isoenzyme activity was elevated (compared with that in well-nourished control subjects consuming normal diets with vitamin D) in 20% of the women postpartum and in 50% of the infants. Five infants developed symptomatic hypocalcemia.
In a double-blind study of supplementary vitamin D (1000 IU/d, or 25 µg/d) administered during the third trimester to Asian women living in London, Maxwell et al (46) showed increased maternal weight gain and a 50% reduction in the numbers of infant classified as growth retarded (born weighing <2500 g at term) with supplementation, which closely approached significance at the 5% level. Infants in the control (nonsupplemented) group also had larger fontanelles, suggesting delayed ossification (46). The infants were followed up to the age of 1 y, with the observers still blinded to the original randomization. The control and supplemented groups diverged increasingly in terms of both weight and length over time, so that by 1 y of age, the infants whose mothers had received supplemental vitamin D during the third trimester were on average 0.4 kg heavier and 1.6 cm longer than infants in the control group (47, 48).
Thus, maternal malnutrition with coexisting vitamin D deficiency can result in metabolic bone disease and disturbed calcium and vitamin D metabolism in neonates. Vitamin D supplementation of malnourished mothers results in improved growth of the fetus and child, both in terms of birth weight and subsequent linear growth during infancy (49–51).
Therefore, if daily vitamin D supplementation during the whole pregnancy can be undertaken, the amount given should be 400 IU/d (10 µg/d). In countries where dairy products are not supplemented with vitamin D, where sunshine exposure is low (in northern countries), or when presentation for antenatal care is delayed, 1000 IU/d (25 µg/d) should be given during the last 3 mo of pregnancy or 100000 IU (2500 µg/d) in one dose at the beginning of the last trimester (52, 53).
FULL-TERM NEWBORNS
Abruptly severed from the placental supply of nutrients after birth, newborns must adapt rapidly to ensure positive calcium balance for normal skeletal growth and development. In healthy, full-term newborns, total and ionized calcium concentrations progressively decrease after birth, so that by the second or third day of life calcium concentrations are often lower than that those found in older infants and children (17). Calcium concentrations usually return to normal by 5–10 d of age. Serum PTH concentrations tend to be low in cord blood but increase within the first 48 h of life in response to the decrease in serum calcium. The latter also induces increased synthesis of 1,25-dihydroxyvitamin D to the 5th day of life in a normal fashion (54–56).
At birth, serum cord osteocalcin concentrations are 2–3 times higher than concentrations in healthy adults and in the pregnant mother; osteocalcin concentrations in the newborn increase significantly between birth and days 1–5 of life and decrease thereafter. A significant, positive correlation between osteocalcin and 1,25-dihydroxyvitamin D is found in cord blood and at day 5 (57).
We showed in a controlled study that maternal vitamin D supplementation dampens the decrease in serum calcium observed in newborns at 4 d of age (58). Furthermore, in both vitamin D supplemented and unsupplemented groups the magnitude of the difference between calcium concentrations in cord blood and those observed at 4 d of age was related to the concentration in cord blood; at any calcium concentration, the decrease was less marked in infants of supplemented mothers. Because calcium intakes during the first week of life are low, even if during this period newborns receive a formula richer in calcium than human milk, the unsupplemented infants could not correct this hypocalcemia. The exact mechanism by which maternal supplementation of vitamin D affects perinatal calcium homeostasis remains speculative. Supplementation may allow ready transfer of a large pool of 25-hydroxyvitamin D to the newborn followed by rapid renal synthesis of 1,25-dihydroxyvitamin D to meet the needs of the newborn: this view is supported by the higher values of circulating 1,25-dihydroxyvitamin D observed at 4 d of age in infants born to supplemented mothers (58).
Longitudinal measurement of 25-hydroxyvitamin D concentrations in the serum of breast-fed infants not receiving vitamin D supplements and who were born to vitamin D–replete mothers suggest that vitamin D stores are depleted within 8 wk of delivery. This finding implies that the vitamin D content of human milk is low (59). The effect of vitamin D supplementation on bone mineralization in extensively breast-fed infants was investigated by Greer et al (60). These investigators showed that infants who received 400 IU vitamin D/d (10 µg/d) had higher bone mineral content and serum 25-hydroxyvitamin D concentrations than did control infants.
In 1963 the American Academy of Pediatrics recommended that the daily intake of vitamin D in full-term infants not be <400 IU (10 µg/d). Although these recommendations are still regarded as appropriate by many pediatricians, this requirement is fulfilled in formula-fed infants but not in breast-fed infants (60).
PREMATURE INFANTS
Early neonatal hypocalcemia affects 75% of preterm infants during the first days of life, principally those born with very low birth weights (<1500 g) (61). The hypocalcemia is usually of short duration and may not result in clinical symptoms in most infants. Early neonatal hypocalcemia may be caused by immaturity of the vitamin D activation pathway, either alone or in combination with other abnormalities, particularly transient hypoparathyroidism, hypercalcitoninemia, and end organ resistance to hormonal effects.
It has been clearly shown that there is an appropriate secretion of PTH in response to this hypocalcemic stimulus (62–64). Serum immunoreactive PTH concentrations increase immediately after birth. Both the intact hormone (amino acids 1–84) and the carboxy terminal fragment follow the same trend, indicating that in preterm infants secretion of the hormone responds physiologically to the hypocalcemic stimulus (Table 3). This is substantiated by a report indicating that the increase in immunoreactive PTH concentrations was blunted when premature infants received calcium by infusion; this calcium load buffered the postnatal depression of serum calcium (64). By day 10, serum concentrations of intact PTH and carboxy terminal PTH returned to euparathyroid values (Table 3).
View this table:
TABLE 3.. Calcium, intact parathyroid hormone [PTH(1–84)], carboxy terminal PTH (cPTH), and vitamin D binding protein (DBP) concentrations in cord serum and at days 1, 2, 5, 10, and 30 after birth in 15 preterm infants1
As in full-term newborns, in preterm newborns both total and free 25-hydroxyvitamin D concentrations in cord blood are lower than those in maternal blood and are correlated with concentrations in the mother (66). Bouillon et al (26) reported a positive correlation between maternal and cord serum concentrations of both total and free 1,25-dihydroxyvitamin D in premature infants. In our experience, however, only concentrations of free 1,25-dihydroxyvitamin D are correlated; the discrepancy between our results and those of Bouillon et al may be due to the vitamin D depletion in our patients (57). Concentrations of cord and maternal blood vitamin D binding protein were also positively correlated in Bouillon et al's study. By analyzing polymorphisms in vitamin D binding protein, Hirsfeld and Lunell (67) excluded the possibility of placental transfer of this protein. Thus, the most likely explanation for this fetomaternal relation is common fetal and maternal factors affecting the synthesis of vitamin D binding protein.
Many reports have clearly shown that in infants born after 28 wk of gestation, activation of vitamin D is operative as early as 24 h after birth (66, 68–72). Vitamin D supplementation just after birth improves vitamin D nutritional status as evidenced by rising plasma 25-hydroxyvitamin D concentrations (Figure 2). Indeed, in France and other European countries where dairy products are not enriched with vitamin D, mean cord concentrations of 25-hydroxyvitamin D are lower than in North America (3). In a controlled study, Koo et al (73) measured plasma 25-hydroxyvitamin D concentrations in preterm infants receiving 200, 400, or 800 IU vitamin D/d (5, 10, or 20 µg/d). They showed that plasma 25-hydroxyvitamin D remained normal for 6 mo while infants received <400 IU vitamin D/d (<10 µg/d).
FIGURE 2. . Mean (±SEM) serum total 25-hydroxyvitamin D [25(OH)D] and 1,25-dihydroxyvitamin D [1,25(OH)2D] concentrations as a function of age in 15 preterm infants (birth weight: 1578 ± 78 g; gestational age: 31.7 ± 0.5 wk) who received 1000 IU (25 µg) vitamin D/d from birth. *,**,***Significantly different from cord serum (one-factor, repeated-measures ANOVA):*P < 0.05, **P < 0.01, ***P < 0.001.
Administration of vitamin D (from 1000 IU/d, or 25 µg/d) also results in an increase in circulating concentrations of total 1,25-dihydroxyvitamin D (Figure 2). By 5 d of age, plasma concentrations of total 1,25-dihydroxyvitamin D in supplemented infants are well above the range observed in reference adolescent groups. This sharp elevation in 1,25-dihydroxyvitamin D is probably linked to hypocalcemia and concomitant elevated PTH concentrations. Substrate concentration is a rate limiting factor in the synthesis of 1,25-dihydroxyvitamin D in the presence of hypocalcemia. Accordingly, a strong positive correlation between serum free and total 25-hydroxyvitamin D and 1,25-dihydroxyvitamin D concentrations was observed during the first 10 d of life over a wide range of 25-hydroxyvitamin D concentrations (65, 68, 70).
Delvin and others (unpublished observations, 1994) showed that 1,25-dihydroxyvitamin D receptors are present in the fetal gut as early as 13 wk of gestation. This observation suggests that intestinal cells in premature infants are likely responsive to the hormone from that point on.
Serum osteocalcin values are high at birth (15 ± 3 µg/L) and maternal and cord serum osteocalcin concentrations are not correlated (57). During the first month of life, serum osteocalcin increases and parallels the changes in 1,25-dihydroxyvitamin D without sustained correlation (57). These results indicate than serum osteocalcin does not reflect changes in serum 1,25-dihydroxyvitamin D but may reflect primarily the overall rate of bone formation or growth at the tissue level.
After the first week of life, plasma 25-hydroxyvitamin D remains constant in preterm infants who received supplemental vitamin D; 1,25-dihydroxyvitamin D concentrations increase up to 30 d (Figure 2 and Table 4), with no further change until the end of the first 3 mo of age (68). Concentrations of 1,25-dihydroxyvitamin D in preterm infants up to 3 mo of age are >2–3 times higher than those seen in older children. During this period, there is no significant correlation between vitamin D metabolite concentrations and serum calcium and phosphorus concentrations and calcium and phosphorus intake. At 30 d of age, there is also no influence of diet (formula or fortified human milk) on serum mineral or vitamin D metabolite indexes (Table 4).
View this table:
TABLE 4.. Effect of nutrition on biochemical indexes of calcium homeostatsis at 30 d of life in 15 preterm infants fed breast milk or preterm formula1
The high concentrations of plasma 1,25-dihydroxyvitamin D beyond the perinatal period (after 1 mo of age) (68) may represent a compensatory effect to ensure calcium and phosphorus absorption from the diet at a time when bone mineralization occurs. Osteopenia is common in premature infants, particularly in those who underwent a prolonged period of parenteral feeding or received a diet low in calcium and phosphorus (European formula or human milk). There is now widespread agreement that a deficiency of mineral substrates and not intake or metabolism of vitamin D is the main etiologic factor in osteopenia in preterm infants (74, 75).
The vitamin D requirements of preterm infants are influenced by the body stores at birth, which in turn are related to the length of gestation and maternal stores. These factors should be taken into consideration when vitamin D supplementation policies are developed in each country. The American Academy of Pediatrics recommends that preterm infants receive 400 IU vitamin D/d (10 µg/d) independent of that contained in low-birth-weight formula (76). The European Society of Paediatric Gastroenterology and Nutrition recommends that when preterm infants are fed human milk, they receive a supplement of 1000 IU (25 µg) vitamin D/d. Formula-fed infants should also be supplemented with vitamin D to achieve the same intake as infants receiving breast milk. The maximum vitamin D content of a formula should not exceed 30 µg/L (1200 IU) (77).
INFANTS OF DIABETIC MOTHERS
Hypocalcemia in infants of diabetic mothers has been the subject of numerous studies and several pathogenic factors have been suggested in this metabolic disorder, including hypoparathyroidism, hyperphosphatemia, hypomagnesemia, and defects in vitamin D metabolism (78). Hypocalcemia in infants of diabetic mothers appears during the very early hours of life and changes little after 24 h of age. This hypocalcemia tends to be more severe than the early neonatal hypocalcemia observed in preterm infants and tends to persist longer (79). No consistent abnormality in vitamin D metabolism has been observed in infants of diabetic mothers, similar to the observation in premature infants with early neonatal hypocalcemia (80). Thus, the pathogenesis of hypocalcemia in infants of diabetic mothers remains unclear. The bone mass of infants of diabetic mothers at birth is significantly higher than that of normal infants with the same weight (81). Because infants of diabetic mothers are hyperinsulinemic, especially the large-for-gestational-age infants, the increase in fetal mineralization could be explained by the known effect of insulin and insulin-like growth factor 1 on bone formation. The increased bone mass of these infants may be responsible for their increased calcium needs and subsequently for the neonatal hypocalcemia.
CONCLUSION
Obstetricians and pediatricians should focus on providing sufficient vitamin D during pregnancy and at birth to term and preterm infants. Additional calcium and phosphorus should also be given early in life, particularly to preterm infants. In countries where dairy products are not routinely supplemented with vitamin D or where sunshine exposure is low due because of geographic location or religious beliefs, vitamin D should be administered at the end of pregnancy and just after birth to newborns.
ACKNOWLEDGMENTS
We thank P Delmas (Inserm 403, Lyon, France) for serum osteocalcin measurements, R Bouillon (Leuven, Belgium) for measurement of vitamin D binding protein, and G Gueris (Lariboisière, Paris) for measurement of serum parathyroid hormone. We are also grateful to Ronald Hagan (Perth, Australia) for reviewing and correcting the manuscript.
REFERENCES