The effect of 1,25(OH)2 vitamin D3 supplementation in premature infants

Effects of oral vitamin D supplementation on linear growth and other health outcomes among children under five years of age

BACKGROUND

Vitamin D is a secosteroid hormone that is important for its role in calcium homeostasis to maintain skeletal health. Linear growth faltering and stunting remain pervasive indicators of poor nutrition status among infants and children under five years of age around the world, and low vitamin D status has been linked to poor growth. However, existing evidence on the effects of vitamin D supplementation on linear growth and other health outcomes among infants and children under five years of age has not been systematically reviewed.

OBJECTIVES

To assess effects of oral vitamin D supplementation on linear growth and other health outcomes among infants and children under five years of age.

SEARCH METHODS

In December 2019, we searched CENTRAL, PubMed, Embase, 14 other electronic databases, and two trials registries. We also searched the reference lists of relevant publications for any relevant trials, and we contacted key organisations and authors to obtain information on relevant ongoing and unpublished trials.

SELECTION CRITERIA

We included randomised controlled trials (RCTs) and quasi-RCTs assessing the effects of oral vitamin D supplementation, with or without other micronutrients, compared to no intervention, placebo, a lower dose of vitamin D, or the same micronutrients alone (and not vitamin D) in infants and children under five years of age who lived in any country.

DATA COLLECTION AND ANALYSIS

We used standard Cochrane methodological procedures.

MAIN RESULTS

Out of 75 studies (187 reports; 12,122 participants) included in the qualitative analysis, 64 studies (169 reports; 10,854 participants) contributed data on our outcomes of interest for meta-analysis. A majority of included studies were conducted in India, USA, and Canada. Two studies reported for-profit funding, two were categorised as receiving mixed funding (non-profit and for-profit), five reported that they received no funding, 26 did not disclose funding sources, and the remaining studies were funded by non-profit funding. Certainty of evidence varied between high and very low across outcomes (all measured at endpoint) for each comparison. Vitamin D supplementation versus placebo or no intervention (31 studies) Compared to placebo or no intervention, vitamin D supplementation (at doses 200 to 2000 IU daily; or up to 300,000 IU bolus at enrolment) may make little to no difference in linear growth (measured length/height in cm) among children under five years of age (mean difference (MD) 0.66, 95% confidence interval (CI) -0.37 to 1.68; 3 studies, 240 participants; low-certainty evidence); probably improves length/height-for-age z-score (L/HAZ) (MD 0.11, 95% CI 0.001 to 0.22; 1 study, 1258 participants; moderate-certainty evidence); and probably makes little to no difference in stunting (risk ratio (RR) 0.90, 95% CI 0.80 to 1.01; 1 study, 1247 participants; moderate-certainty evidence). In terms of adverse events, vitamin D supplementation results in little to no difference in developing hypercalciuria compared to placebo (RR 2.03, 95% CI 0.28 to 14.67; 2 studies, 68 participants; high-certainty evidence). It is uncertain whether vitamin D supplementation impacts the development of hypercalcaemia as the certainty of evidence was very low (RR 0.82, 95% CI 0.35 to 1.90; 2 studies, 367 participants). Vitamin D supplementation (higher dose) versus vitamin D (lower dose) (34 studies) Compared to a lower dose of vitamin D (100 to 1000 IU daily; or up to 300,000 IU bolus at enrolment), higher-dose vitamin D supplementation (200 to 6000 IU daily; or up to 600,000 IU bolus at enrolment) may have little to no effect on linear growth, but we are uncertain about this result (MD 1.00, 95% CI -2.22 to 0.21; 5 studies, 283 participants), and it may make little to no difference in L/HAZ (MD 0.40, 95% CI -0.06 to 0.86; 2 studies, 105 participants; low-certainty evidence). No studies evaluated stunting. As regards adverse events, higher-dose vitamin D supplementation may make little to no difference in developing hypercalciuria (RR 1.16, 95% CI 1.00 to 1.35; 6 studies, 554 participants; low-certainty evidence) or in hypercalcaemia (RR 1.39, 95% CI 0.89 to 2.18; 5 studies, 986 participants; low-certainty evidence) compared to lower-dose vitamin D supplementation. Vitamin D supplementation (higher dose) + micronutrient(s) versus vitamin D (lower dose) + micronutrient(s) (9 studies) Supplementation with a higher dose of vitamin D (400 to 2000 IU daily, or up to 300,000 IU bolus at enrolment) plus micronutrients, compared to a lower dose (200 to 2000 IU daily, or up to 90,000 IU bolus at enrolment) of vitamin D with the same micronutrients, probably makes little to no difference in linear growth (MD 0.60, 95% CI -3.33 to 4.53; 1 study, 25 participants; moderate-certainty evidence). No studies evaluated L/HAZ or stunting. In terms of adverse events, higher-dose vitamin D supplementation with micronutrients, compared to lower-dose vitamin D with the same micronutrients, may make little to no difference in developing hypercalciuria (RR 1.00, 95% CI 0.06 to 15.48; 1 study, 86 participants; low-certainty evidence) and probably makes little to no difference in developing hypercalcaemia (RR 1.00, 95% CI 0.90, 1.11; 2 studies, 126 participants; moderate-certainty evidence). Four studies measured hyperphosphataemia and three studies measured kidney stones, but they reported no occurrences and therefore were not included in the comparison for these outcomes.

AUTHORS' CONCLUSIONS

Evidence suggests that oral vitamin D supplementation may result in little to no difference in linear growth, stunting, hypercalciuria, or hypercalcaemia, compared to placebo or no intervention, but may result in a slight increase in length/height-for-age z-score (L/HAZ). Additionally, evidence suggests that compared to lower doses of vitamin D, with or without micronutrients, vitamin D supplementation may result in little to no difference in linear growth, L/HAZ, stunting, hypercalciuria, or hypercalcaemia. Small sample sizes, substantial heterogeneity in terms of population and intervention parameters, and high risk of bias across many of the included studies limit our ability to confirm with any certainty the effects of vitamin D on our outcomes. Larger, well-designed studies of long duration (several months to years) are recommended to confirm whether or not oral vitamin D supplementation may impact linear growth in children under five years of age, among both those who are healthy and those with underlying infectious or non-communicable health conditions.

Prenatal factors associated with the neonatal line thickness in human deciduous incisors

The neonatal line (NNL) is used to distinguish developmental events observed in enamel which occurred before and after birth. However, there are few studies reporting relationship between the characteristics of the NNL and factors affecting prenatal conditions. The aim of the study was to determine prenatal factors that may influence the NNL thickness in human deciduous teeth. The material consisted of longitudinal ground sections of 60 modern human deciduous incisors obtained from full-term healthy children with reported birth histories and prenatal factors. All teeth were sectioned in the labio-lingual plane using diamond blade (Buechler IsoMet 1000). Final specimens were observed using scanning electron microscopy at magnifications 320×. For each tooth, linear measurements of the NNL thickness were taken on its labial surface at the three levels from the cemento-enamel junction. The difference in the neonatal line thickness between tooth types and between males and females was statistically significant. A multiple regression analyses confirmed influence of two variables on the NNL thickness standardised on tooth type and the children's sex (z-score values). These variables are the taking of an antispasmodic medicine by the mother during pregnancy and the season of the child's birth. These two variables together explain nearly 17% of the variability of the NNL. Children of mothers taking a spasmolytic medicine during pregnancy were characterised by a thinner NNL compared with children whose mothers did not take such medication. Children born in summer and spring had a thinner NNL than children born in winter. These results indicate that the prenatal environment significantly contributes to the thickness of the NNL influencing the pace of reaching the post-delivery homeostasis by the newborn's organism.

Calcium nutrition and metabolism during infancy

Calcium is a vital mineral for the developing newborn infant. This review discusses perinatal and neonatal calcium metabolism, with an emphasis on enteral calcium absorption and the nutritional factors affecting calcium bioavailability including the three major endocrine hormones involved in calcium metabolism: parathyroid hormone, vitamin D, and calcitonin. The placenta transports calcium to the fetus throughout pregnancy, with the largest amount of fetal calcium accumulation occurring in the third trimester. At birth, the newborn transitions to intestinal absorption to meet the body's calcium needs. Most calcium is absorbed by paracellular passive diffusion in the small intestine. Calcium intestinal absorption is affected by the type and amount of calcium ingested. It is also affected by the amount of intestinal calcium that is bound to dietary fats and proteins. One major consequence of decreased calcium absorption is metabolic bone disease in which there is a failure of complete mineralization of the bone osteoid.

Neonatal hypocalcemia: to treat or not to treat? (A review)

The various definitions of neonatal hypocalcemia (NHC) are critically examined. The authors review the known and less known complications of NHC and of its treatment. They emphasize the need for a definition of NHC in terms of serum ionized rather than total calcium concentrations, the rationale for treating symptomatic NHC, as well as the need for further research in infants with asymptomatic NHC.

Intact parathyroid hormone levels during pregnancy, in healthy term neonates and in hypocalcemic preterm infants

We measured parathyroid hormone levels in pregnant and nonpregnant women and at 1, 2 and 5 days of life in healthy term neonates and in hypocalcemic preterm infants using a new immunoradiometric assay which measures only biologically active intact parathyroid hormone and by a mid-molecule parathyroid hormone radioimmunoassay. During pregnancy intact and mid-molecule parathyroid hormone levels did not show any modification and were not different from parathyroid hormone levels of nonpregnant age-matched controls. Serum calcium and phosphorus levels did not vary during each trimester of pregnancy. In cord serum intact and mid-molecule parathyroid hormone values were low in both term and preterm infants. In term neonates intact and mid-molecule parathyroid hormone levels peaked on day 1; in preterm infants intact parathyroid hormone levels peaked on day 1 while mid-molecule parathyroid hormone values peaked on day 2. Intact parathyroid hormone levels showed a more marked increase in preterm (19-fold) than in term neonates (7.5-fold) on day 1. Our data do not confirm the previously reported "physiologic" hyperparathyroidism in pregnancy. Moreover we found a normal parathyroid gland responsiveness to decreasing serum calcium levels in the first days of life in term and preterm infants. Our results suggest that measurement of intact parathyroid hormone 1-84 by immunoradiometric assay in the first days of life is a more sensitive index of parathyroid gland secretory function than the measurement of middle or carboxyl-terminal parathyroid hormone fragments allowing the detection of the dynamic changes of parathyroid hormone which occur in hypocalcemic preterm infants.

Plasma ionized calcium, phosphate and magnesium in preterm and small for gestational age infants

Blood concentrations of ionized calcium, phosphate and magnesium were determined on days 1, 3 and 5 in 32 preterm infants, appropriate for gestational age, and 25 small for gestational age infants. The results were compared with those of a reference group of 31 fullterm newborns. Preterm infants had lower mean ionized calcium levels than the reference population. Ionized calcium levels were positively correlated with gestational age and postnatal age. The sickest infants had the lowest calcium levels. Intrauterine growth retardation did not seem to influence the calcium levels. Small for gestational age infants had lower phosphate values than the reference group and the values correlated with the degree of growth retardation the first day. Magnesium concentrations increased with time and did not differ significantly between the study groups and the reference group. Magnesium values did not correlate with calcium or phosphate values.

Calcium homeostasis in premature infants and treatment of early hypocalcaemia by 1,25-dihydroxycholecalciferol

We studied calcium homeostasis and the serum calcium response to oral 1,25-dihydroxycholecalciferol [1,25 (OH)2D3] at a low pharmacological dosage of 0.1 microgram/kg daily in 14 early hypocalcaemic asymptomatic neonates. Seven hypocalcaemic neonates were not treated. In hypocalcaemic neonates serum PTH levels were normal, the urinary C-AMP response after PTH stimulation was poor and plasma 1,25 (OH)2D3 was low. Treatment with 1,25(OH)2D3 resulted in a rapid increase of serum calcium. The increase was more rapid in neonates treated with 1,25(OH)2D3 than in untreated subjects. A similar result was obtained in one of a pair of identical twins. These results suggest that a low dose of 1,25(OH)2D3 is effective in treating neonatal hypocalcaemia. However, the response was delayed for 48 h. The reason for this delay is not clear.

Development of vitamin D3 25-hydroxylase activity in rat liver microsomes

The ontogeny of vitamin D3 25-hydroxylase activity has been determined in liver microsomes of rat fetuses and neonates. Production of 25-hydroxyvitamin D3 was low (0.11 pmol/g liver/h) 3 days prior to birth. Production rates were 1.2, 2.2, 1.8, and 2.8 pmol/g liver/h on Day 0, Day 2, Day 7, and Day 15, respectively. 25-Hydroxyvitamin D3 production in neonates increased sixfold from Day 15 to Day 22 to a value twice that of the mothers (17.6 pmol/g liver/h compared with 7.3 pmol/g liver/h). Activity in the maternal microsomes was constant (0.22 to 0.30 pmol/mg protein/h) except for the day of parturition (0.54 pmol/mg protein/h) and Day 22 postpartum (0.44 pmol/mg protein/h). A cytosolic factor, present as early as 3 days prior to birth, was required for vitamin D3 25-hydroxylase activity in the fetuses and stimulated the 25-hydroxylase reaction (up to 2.5-fold) in neonates and mothers. The ability of cytosol to prevent degradation of vitamin D3 was also present in the fetal stage. These data suggest that microsomal vitamin D3 25-hydroxylase activity in rat liver microsomes develops slowly and reaches full activity near the weaning stage. Since the cytosolic factor(s) is/are present in the fetal stage, the limiting component in the maturation of vitamin D3 25-hydroxylase activity in liver microsomes is the development of the cytochrome P-450 vitamin D3 25-hydroxylase.

Circulating immunoreactive parathyroid hormone levels in premature infants and the response to calcium therapy

Serum immunoreactive parathyroid hormone (iPTH), calcium, magnesium, and phosphorus levels were measured in 13 premature infants during the first 96 hours of life. Hypocalcemia at 12-24 hours of age was associated with a markedly elevated mean serum iPTH level. Six of the hypocalcemic infants received a continuous infusion of calcium while seven were not treated. In the untreated infants, the mean serum calcium remained in the hypocalcemic range while the serum iPTH progressively increased. By contrast, the mean serum calcium in the treated infants increased to 2.35 mmol/l at 96 hours of age and was accompanied by a decline in serum iPTH. At 72 and 96 hours, the mean serum iPTH was twofold greater in the untreated than in the treated infants. The results indicate that the parathyroid glands of premature infants respond to calcium signals and that a factor(s), other than parathyroid insufficiency, plays an etiologic role in the hypocalcemia of prematurity.

Seasonal variation in serum-25-OH-D3 in mothers and newborn infants in northern finland

Two groups, each consisting of twenty Finnish mother-neonate pairs and ten non-pregnant controls were studied for serum calcium, serum phosphate, serum alkaline phosphatase, parathormone index (PTHind) and S-25-OH-D3. The first series was collected in winter and the other in summer. The serum samples were taken on the third day after delivery. The concentrations of S-25-OH-D3 were significantly lower in the mothers than in the non-pregnant controls in winter, but the difference was not significant in summer. The concentrations of S-25-OH-D3 in the serum of the mothers were similarly significantly lower in winter than in summer, ten mothers exhibiting a value below the detection line in winter, but only two in summer. The concentrations of S-25-OH-D3 in the mothers and their newborn infants showed a close relationship, but when extremely low values existed in the mothers, the infant concentrations were slightly higher. The seasonal variation in S-25-OH-D3 was also significant in the neonates. Although calcium was decreased and alkaline phosphatase elevated when compared with the non-pregnant controls in the mothers in both groups, these values showed no seasonal variation, and the mean levels of serum calcium, phosphate, alkaline phosphatase and PTHind in the neonates also remained unaltered between the two groups. The results indicate that additional vitamin D should be supplied during pregnancy in the winter months at this latitude.

Do breast and bottle fed babies require vitamin supplements?

In the present stage of knowledge, the following approach to vitamin supplementation seems reasonable. Breast-fed full-term infants should be supplemented with vitamin D (400 I.U./day). Supplementation of bottle-fed infants will depend upon the vitamin content of the formula which is being used. Pre-term and LBW infants may need higher amounts of vitamin D (800-1 000 I.U./day) and should be supplemented with vitamin E (1-2 I.U./day), C (50 mg/day), B6 (30-60 mcg/day) and folic acid (50-100 mcg/day) when they are being fed formulas with high amounts of PUFA and proteins or with iron. Multivitamin supplementation appears to be acceptable for infants of very low birth weight (less than 1 500 g) until they reach a body weight of at least 2 000 grams or a caloric intake of 300 kcal/day. Finally, the present routine of administering 0.5-1.0 mg of vitamin K at birth to all infants should be continued.

Tanning, protection against sunburn and vitamin D formation with a UV-A 'sun-bed'

There are many types of sun-beds, sun-benches and sun-panels containing fluorescent tubes which, because of their predominantly UV-A emission, are advertised to the public as a means of obtaining a tan without sunburn. This study reports the effects of a sun-bed on skin colour, on the protection afforded against sunburn, and on vitamin D formation. Side-effects are also recorded. It was shown that the sun-bed emits mainly UV-A but very little UV-B and some tanning occurred in most subjects. However, no correlation was observed between the subjects' stated ability to tan and the degree of pigmentation achieved at the end of the treatment. Most subjects also had itching and erythema, and three had polymorphic light eruption. Although very little UV-B irradiation was present, a significant increase in serum levels of 25-hydroxyvitamin D occurred, and possible explanations of this surprising finding are discussed. While the sun-bed proved popular with the subjects, only a modest tan was achieved and the incidence of side-effects appeared to limit the value of this type of appliance, especially with regard to the prevention of vitamin D deficiency.

Critical vitamins for low birthweight infants

All vitamins are "critical" by definition for the growing infant. However, some of them are particularly relevant to the preterm or low birthweight (LBW) infant in whom a deficiency can occur more frequently than in a full-term newborn. In LBW infants vitamin deficiency may develop due to (1) low body stores at birth, (2) low intake, (3) limited absorption, (4) increased need or utilization, (5) presence of certain clinical conditions. The first reason concerns all lipid-soluble vitamins, and particularly vitamin E and K, which cross the placenta with some difficulty. Among hydrosoluble vitamins, cord-blood levels of vitamin B6 have been shown to be abnormally low in preterm infants. Low intake can occur because of low vitamin levels in milk or because of delayed and/or insufficient feeding. Limited intestinal absorption of vitamins in LBW infants has only been demonstrated for vitamin E and folic acid. The rapid post-natal growth may lead to increased vitamin utilization. In some clinical conditions particularly high intakes of certain vitamins are indicated. In our opinion, the really "critical" vitamins in LBW infants are vitamin D, E, K and folic acid, for which routine supplementation can be recommended, and possibly vitamins C and B6 under special circumstances.

Effect on serum calcium of a alpha-hydroxy-vitamin D3 supplementation in infants of low birth weight, infants with perinatal asphyxia, and infants of diabetic mothers

Thirty infants of low birth weight, 35 infants with perinatal asphyxia, and 16 infants of diabetic mothers were investigated for early neonatal hypocalcaemia. The infants were randomized into a group prophylactically given 1 alpha-hydroxy-vitamin D3, 0.05 or 0.10 micrograms/kg i.v. on the first 3 days of life, and an untreated control group. In infants of low birth weight and infants of diabetic mothers there were no differences in serum ion-Ca concentrations on days 2, 3, 5, and 7 between the treated and untreated groups. In infants with perinatal asphyxia, however, serum ion-Ca concentrations on days 5 and 7 were significantly higher in the treated than in the untreated group, while on days 2 and 3 the differences were not statistically significant. The hypocalcaemia in asphyctic infants was not correlated to bicarbonate treatment, but infants with severe signs of asphyxia had lower serum ion-Ca concentrations than infants with only mild or no signs. Hypocalcaemia in asphyctic infants might be explained by a decreased concentration of 1 alpha, 25-dihydroxy-vitamin D3 following reduced 1 alpha-hydroxylation in the kidney as a consequence of anoxia during perinatal asphyxia.

Vitamin D metabolism in preterm infants: serum calcitriol values during the first five days of life

To ascertain the activity of the vitamin D biosynthetic pathway, the serum concentration of 1,25-dihydroxyvitamin D (calcitriol) was measured in 16 preterm infants (32 to 37 weeks of gestation) at 1 to 2 and 120 hours of age. Half of the subjects received a daily oral supplement of 2,100 IU of vitamin D3 during the five-day study period. In the first two hours of life, all subjects were hypocalcemic (8.2 +/- 0.2 mg/dl) and 14 subjects had low concentrations of 25-hydroxyvitamin D (calcidiol, 8 +/- 1 ng/ml). The latter finding probably reflects a mild degree of vitamin D deficiency in the mothers of our subjects. Calcitriol concentrations (42 +/- 3 pg/ml) were comparable to those of older children. At 120 hours of age, the control group had no significant change in calcitriol values, whereas the group supplemented with D3 had a more than threefold increase. There was a positive correlation between the circulating concentrations of calcidiol and calcitriol over the period of the study. The data show that, after 32 weeks of gestation, renal 25-hydroxyvitamin D-1 alpha-hydroxylase activity is present, with the rate of calcitriol synthesis being apparently substrate limited. Early neonatal hypocalcemia is therefore unlikely to be caused by an impairment of vitamin D activation.

The vitamin D system: a view from basic science to the clinic

Vitamin D produced in the skin and absorbed in the small intestine must be modified metabolically before it can function. It is ultimately converted to a hormone in the kidney that stimulates intestine, bone and kidney to mobilize calcium and phosphorus. This results in normal bone development and normal neuromuscular function. The vitamin D hormone appears to act by a nuclear mechanism to facilitate a target organ response. Finally the vitamin D hormone is produced in response to the need for calcium and phosphorus. The calcium need is interpreted by the parathyroid gland that in turn secretes parathyroid hormone. The parathyroid hormone stimulates production of the vitamin D hormone. This constitutes the vitamin D endocrine system that plays an important role not only in calcium homeostasis but in phosphate homeostasis and in calcium economy of the body. A number of disease states including hypoparathyroidism, pseudohypoparathyroidism, renal osteodystrophy, certain types of vitamin D-resistant rickets and osteoporosis can in part be related to disturbance in the vitamin D endocrine system. Thus measurement of the vitamin D hormone and its precursor will be of great value in diagnosis of metabolic bone disease and most importantly, the availability of new vitamin D compounds will play an important role in the treatment of these bone diseases.

Plasma active vitamin D concentration in low birthweight infants with rickets and its response to vitamin D treatment

Impairment of 25-hydroxylation may be a cause of rickets in infants of low birthweight, weighing between 2000 and 2500 g. In addition there may be impairment of 1 alpha-hydroxylation in infants weighing less than 2000 g. Our data show that a supplementary dose of vitamin D2 of at least 500 IU daily is a reasonable regimen for infants who weighed between 2000 and 2500 g at birth. However for infants who weighed less than 2000 g with a gestation of under 38 weeks, administration of 1 alpha-OHD3 may be more effective in preventing rickets.

Plasma 25-hydroxyvitamin D concentrations in preterm infants receiving oral vitamin D supplements

Plasma 25-hydroxy-cholecalciferol (25-OHD) was measured serially in two groups of preterm infants receiving either 400 IU or 1000 IU oral vitamin D3 daily. All the babies were able to absorb and hydroxylate vitamin D3 adequately by 36 weeks' gestational age. The higher daily supplement had no advantage over the lower dose.

1,25-Dihydroxyvitamin D3 in childhood hepatic osteodystrophy

Osteodystrophy frequently accompanies severe childhood hepatobiliary disease. Proposed causes include malabsorption of vitamin D and calcium, and diminished 25-hydroxylation of vitamin D. Two children, ages 23 and 35 months, with radiographic and biochemical evidence of rickets with extrahepatic biliary atresia, were treated with 1,25-dihydroxyvitamin D3. The minimal effective therapeutic dose and efficacy of 1,25-(OH)2D3 in the treatment of rickets associated with severe childhood hepatic disease were determined. Oral 1,25-(OH)2D3 was ineffective at doses of 0.10 microgram/kg/day. Parenteral doses of 0.20 microgram/kg/day effectively produced radiographic, bone mineral (photon absorptiometric), and biochemical evidence of healing. The need for four times the physiologic dose of 1,25-(OH)2D3 by the parenteral route suggested enhanced catabolism of, or end-organ resistance to, 1,25-(OH)2D3 in our patients with severe cholestatic liver disease treated with phenobarbital.

Serum concentrations of 25-hydroxyvitamin D in rickets of extremely premature infants

Nine premature infants developed radiographic and biochemical rickets at a mean +/- SD of 12.6 +/- 2.8 weeks of age. Serum 25-hydroxyvitamin D concentrations were all low, with a mean of less than 3.6 +/- 2.1 ng/ml. The mean average daily intake of vitamin D since birth had been 300 +/- 181 IU, and the mean average daily intake during the week of diagnosis was 587 +/- 313 IU. All of the infants were extremely premature (mean weight 948 +/- 153 gm, mean gestation 27.7 +/- 1.1 weeks), and were being fed either a low-calcium "human milk-like" formula or a soy formula. It is postulated that low-calcium intake may have increased 25-OHD utilization in the face of a decreased ability of the extremely premature infant to produce 25-OHD. Because of multiple factors leading to both decreased production and possible increased utilization of 25-OHD, such infants have an increased requirement for vitamin D to maintain normal serum 25-OHD concentrations, and daily intakes of at least 400 IU vitamin D orally must be assured. Serum 25-OHD measurements and radiographs may be important in following infants at risk.