The fetal and infant periods of childhood development are considered to be pivotal in optimizing adolescent and adult health outcomes. Over the past three decades, since Barker first published evidence that early life events influence outcomes in later life (the ‘Barker hypothesis’),1,2 there has been growing acceptance of the importance of optimal fetal and infant growth in reducing the risk of metabolic diseases and possibly osteoporosis in adults. Furthermore, nutritional factors during the infant and early childhood periods play important roles not only in ensuring optimal bone growth and development during these periods but also in preventing long-term complications, the nutritional factors of importance being vitamin D and dietary calcium.3,4
This review will discuss the effects of vitamin D on bone development during the infant and early childhood periods and the possible long-term consequences of derangements in maternal vitamin D status. Aspects of this topic have recently been reviewed in several articles.5,6
Vitamin D during fetal and infant development
Vitamin D effects on fetal and neonatal development
The role of vitamin D in fetal bone development and calcium homeostasis during pregnancy has been extensively investigated, particularly in animal models. Despite the increased maternal demands for calcium during pregnancy, it is apparent not only from these animal models but also from human studies that vitamin D plays a negligible role in both maternal and fetal calcium homeostasis.7 It is only after delivery, when the infant becomes dependent on intestinal calcium absorption to maintain normal calcium homeostasis, that ensuring vitamin D sufficiency becomes a priority.8,9 Evidence that neither the vitamin D receptor nor 1,25-dihydroxyvitamin D is required for calcium homeostasis and bone development in the fetus was provided by the finding that infants born with abnormalities of the vitamin D receptor (vitamin D-dependent rickets type IIA) or with vitamin D dependency rickets [1α-hydroxylase (CYP27B1) deficiency (vitamin D-dependent rickets type 1A)] are normal at delivery and only develop hypocalcaemia and rickets several months post delivery. In a series of patients with vitamin D-dependent rickets type 1A, clinical presentation from hypocalcaemia, rickets or delays in motor milestones did not occur until after 5 months of age.10 One of the subjects, who was asymptomatic at the time, was examined at one month of age because of a family history of the disease and at that stage serum calcium, phosphorus, and alkaline phosphatase concentrations were within the normal range despite the serum 1,25-dihydroxyvitamin D [1,25-(OH)2D] concentration being lower than one would expect.
Despite the very convincing evidence outlined above that fetal bone development is relatively uninfluenced by the vitamin D status of the fetus, there are a couple of studies suggesting that maternal vitamin D status might influence fetal bone development.11,12 A Japanese study which investigated the association between vitamin D deficiency and the presence of craniotabes in the neonate at birth found that the incidence of craniotabes was associated with the month of birth (i.e. it was more common in infants born in late winter and spring),11 and the authors attributed this finding to vitamin D deficiency in the late stages of pregnancy. Approximately 27% of those infants with craniotabes at birth still had craniotabes at one month of age, and a similar number of those infants with craniotabes at birth had early features of rickets on radiography at 1 month, and one-third had 25-hydroxyvitamin D (25O-HD) concentrations < 10 ng/ml (25 nmol/l) (all breastfed). Whether or not maternal vitamin D deficiency can be held responsible for the development of craniotabes in these infants is difficult to assess as cord blood 25O-HD levels were not obtained and biochemistry on control subjects at 1 month of age was not available. However, there is evidence that maternal vitamin D deficiency might influence fetal skull mineralization; in a double-blind placebo-controlled study of pregnant south Asian women in the UK, vitamin D supplementation (1000 IU/day) during the last trimester of pregnancy was found to reduce the size of the anterior fontanelle at birth.12
There is also evidence that post-delivery calcium homeostasis might be impaired in neonates born to vitamin D-deficient mothers,13,14 resulting in a higher prevalence of late neonatal hypocalcaemia; however, a number of these studies are observational and lack adequate control groups. In a prospective study of infant vitamin D supplementation, Zeghoud et al.15 showed an inverse relationship between the incidence of neonatal hypocalcaemia at 3–6 days after birth and neonatal 25O-HD concentrations. Similarly, maternal vitamin D supplementation of pregnant women with a high prevalence of vitamin D deficiency has been shown to reduce the incidence of neonatal hypocalcaemia.12,16 Although rarely reported, cases of so-called congenital rickets have been described occurring in the first month after birth.17,18 Typically, these neonates are born to mothers who themselves have frank osteomalacia.19,20
Although it is suggested that maternal calcium homeostasis during pregnancy is independent of vitamin D status,21 there is evidence that maternal vitamin D deficiency is associated with lower maternal calcium and higher parathyroid hormore (PTH) concentrations at parturition, which are similarly reflected in the neonate (cord blood).22
The influence of maternal vitamin D status on fetal growth and size at birth is inconclusive. In a study of intrauterine fetal femoral growth measured by high-resolution ultrasound, poor maternal vitamin D status was associated with greater fetal femoral metaphyseal cross-sectional area and splaying index but vitamin D status did not influence femoral length at either 19 or 34 weeks' gestation.23 The lack of an effect of maternal vitamin D status on body size at birth is corroborated by the studies from Southampton, UK, in which no influence of maternal vitamin D status was found on birth size or size of the offspring at 9 months and 9 years.24 A similar lack of effect of vitamin D supplementation on birthweight was found in a small placebo-controlled trial conducted in France.25 An interesting study from Australia suggested that birthweight was not influenced by maternal vitamin D status in the first trimester, but that vitamin D status in the last trimester did influence neonatal tibial length and birthweight even after adjusting for gestational age.26 Although no significant differences in the various anthropometric measurements of the offspring were found in a randomized controlled trial of vitamin D in the last trimester of pregnancy in south Asian women in the UK, the incidence of small-for-gestational-age infants was reduced by almost half from 29% to 15% in the supplemented mothers; however, this difference was not statistically significant, possibly due to the small sample sizes.12 This finding is supported by a more recent study which found that the odds ratio (OR) for small for gestational age (SGA) was 7.5 if white mothers had 25O-HD levels < 37.5 nmol/l early in pregnancy.27 This association was not found in black mothers. A number of studies have found correlations between maternal vitamin D status and birthweight. In a study conducted over 30 years ago, vitamin D supplementation of pregnant women in India was found to increase neonatal birthweight,28 while a US study found that birthweight was associated with maternal vitamin D intake after controlling for energy and other nutrient intakes.29 There was a 60 g difference in birthweight between infants born to mothers who had estimated vitamin D intakes above 200 IU/day and those born to mothers whose intakes were below 200 IU/day. In an Australian study looking at the factors influencing maternal vitamin D status, the authors found that babies born to vitamin D-deficient mothers (25O-HD < 25 nmol/l) were significantly light (−151 g), even after adjusting for gestational age, maternal age and ethnicity.30 A recent study conducted among different ethnic groups in the Netherlands found that maternal vitamin D deficiency (< 30 nmol/l) in the early part of the second trimester of pregnancy was associated with lower neonatal birthweights (−114.4 g) and a higher risk of being SGA (OR 2.4).31 Of interest was the finding that those infants born to vitamin D-deficient mothers had accelerated weight and length gain during the first year of life and were heavier than infants in the vitamin D-sufficient groups at one year.
Long-term consequences of maternal vitamin D status on growth and bone mass
Although the evidence that maternal vitamin D deficiency influences neonatal calcium homeostasis and is a factor in the pathogenesis of infantile vitamin D deficiency and rickets is substantial, the long-term consequences of maternal vitamin D deficiency on the bone mass and growth of the child are unclear. In a longitudinal study from the UK, maternal vitamin D status in late pregnancy was associated with whole-body bone mass in their children at 9 years of age,32 and this relationship was independent of the children's body stature. These results are supported by another UK study that found a relationship between maternal ultraviolet (UV) light exposure during the last trimester and bone area in the offspring at 9.9 years.33 These findings need to be confirmed in other longitudinal studies and in trials of maternal vitamin D supplementation, as the implications of the findings might be of considerable importance in reducing the risk of osteoporosis in later life. These findings are of interest too, as there is evidence that maternal vitamin D deficiency may be associated with fetal skeletal length or bone mineral content,7 thus suggesting that the effects of maternal vitamin D deficiency not only influence neonatal bone development but also have long-term post-natal consequences. This hypothesis is supported by a recently published study from Finland, which documented lower tibial bone mineral content and cross-sectional area using peripheral quantitative computed tomography (pQCT) in neonates born to mothers with 25O-HD concentrations below 42.6 nmol/l.34 Although by 14 months of age there were no differences in bone mineral content between infants who had been in the high and low vitamin D groups at birth,35 the infants in the high vitamin D group nevertheless had greater tibial cross-sectional area than those in the low vitamin D group, despite the vitamin D status of both groups being similar at that age.
A study conducted in Sheffield, UK, was unable to show any relationship between umbilical cord 25O-HD concentrations and infant weight and height at either 3 or 6 months of age, despite over 70% of the cord blood 25O-HD concentrations being < 20 nmol/l.36 These findings are contrary to those of Brooke et al.37 who randomly supplemented south Asian pregnant women with vitamin D. Although there were no differences in birthweights between the groups, the infants of the supplemented group had better growth to the point that their weights and heights were significantly better at nine and 12 months than those of the infants of the nonsupplemented mothers. It is possible that these results could be explained as being due to a number of post-natal factors, including differing nutrient intakes between the two groups, rather than maternal vitamin D status, as these factors were not assessed. It has also been shown that a lack of vitamin D supplementation of the infant might reduce postnatal growth in populations at high risk of vitamin D deficiency.
Vitamin D and the breastfed infant
Infancy is one of the periods during the course of life when the child is particularly at risk of vitamin D deficiency. Exclusively breastfed infants are especially at risk because of their dependence on breastmilk for their supply of vitamin D, particularly as parents are being cautioned about the long-term detrimental skin consequences of UV radiation from sunlight38 and are thus limiting infant skin exposure to sunlight. At birth, the newborn's vitamin D status is dependent on the vitamin D status of the mother; it is only 25O-HD that crosses the placental barrier and provides the fetus and neonate with a reserve until exogenous supplies are established post delivery. As circulating 25O-HD has a half-life of approximately three weeks, 25O-HD concentrations in an infant born to replete mothers are likely to be adequate for some 6 weeks post natally, but after that a continuous supply of vitamin D is required to prevent the development of vitamin D deficiency. An infant's vitamin D may be derived from supplements, breastmilk or fortified infant milk formulas, or UV radiation of the skin.
Over the past few decades, there has been renewed interest in the role of breastmilk in maintaining the vitamin D status of exclusively breastfed infants rather than relying on vitamin D supplementation of the infant, particularly as vitamin D supplementation of infants is seen to be going against the concept of exclusive breastfeeding which implies that breastmilk contains all the necessary nutrients for sustained and optimal growth during the first 6 months of life. Unlike the situation with regard to the fetus, with 25O-HD being the major vitamin D metabolite crossing the placenta and maintaining fetal vitamin D sufficiency, it appears that the parent vitamin itself, vitamin D, is the major form of vitamin D that is present in breastmilk and is responsible for maintaining the vitamin D status of the infant.7 Typically, breastmilk from unsupplemented mothers has an anti-rachitic activity of ≈ 30–50 IU/l,39 which translates into an infant receiving approximately 25–40 IU daily from breastmilk, an amount that is < 10% of the estimated adequate intake (AI) for infants.40 Several studies in the 1980s suggested that a maternal vitamin D supplement of between 1000 and 2000 IU/day was required to prevent vitamin D deficiency in the breastfed infant.41,42 More recent studies have confirmed that intakes of vitamin D (2000 IU/day) in lactating mothers can achieve mean 25O-HD in nursing infants of > 20 ng/ml (50 nmol/l),39 whereas intakes of 6400 IU/day in mothers result in mean 25O-HD in infants of > 40 ng/ml (100 nmol/l).43 Although the latter high dosage of vitamin D was not shown to result in any adverse effects in the mothers or their neonates, there is consensus that doses in this range should not be used until further large-scale studies have verified their safety. Currently, the Institute of Medicine has set a tolerable upper intake level for vitamin D for pregnant and lactating mothers at 4000 IU/day.40
With vitamin D3 becoming more readily available as the form of vitamin D in supplements and fortificants, a number of studies have compared the efficacy of D2 with that of D3 in different situations with conflicting results; however, there have been few studies conducted in lactating mothers and their infants. Nevertheless, it has been suggested that supplements of vitamin D3 are more effective than those of D2 in increasing infant 25O-HD concentrations.44
25-Hydroxyvitamin D values and the assessment of vitamin D sufficiency
There has been and continues to be much debate about how to measure vitamin D sufficiency or deficiency. This discussion is complicated by the realization that vitamin D may have many beneficial effects besides those associated with bone health and calcium homeostasis.45 The Institute of Medicine has recently reviewed the dietary requirements for vitamin D and calcium throughout the life course.40 In its deliberations, the Committee also considered if there were sufficient data for it to be able to provide recommendations on the vitamin D intakes needed to optimize the non-skeletal actions of vitamin D. It concluded that currently there are insufficient reliable data for it to be able to draw any firm conclusions about the requirements for the non-classical actions of vitamin D. It did, however, support the use of serum 25O-HD concentrations as the measure of vitamin D exposure (representing both dietary intake and endogenous synthesis in the skin). The committee equated a serum 25O-HD level of 50 nmol/l (20 ng/ml) with the recommended dietary allowance (RDA) for vitamin D and a 25O-HD concentration of 40 nmol/l (16 ng/ml) with the estimated average requirement (EAR). As a consequence of these cut-offs, it is suggested that the following 25O-HD concentrations should be used to define risks of inadequacy and deficiency: sufficiency > 50 nmol/l (20 ng/ml); increased risk of inadequacy < 40 nmol/l (16 ng/ml); and increased risk of deficiency < 30 nmol/l (12 ng/ml).46 These values are very similar to those used extensively in current paediatric literature,47,48 although a concentration of ≥ 75 nmol/l (30 ng/ml) has been used by some researchers to define sufficiency.49 It is generally accepted that vitamin D deficiency rickets occurs at 25O-HD concentrations < 25 nmol/l50 and more especially at values < 12.5 nmol/l, in the face of adequate calcium intakes.3
The situation is very different in children with dietary calcium deficiency rickets, as in these children the evidence points to the primary underlying abnormality being a dietary calcium intake insufficient to meet the requirements of the growing skeleton. In these children the circulating vitamin D profile is thus characterized by markedly elevated 1,25-dihdroxyvitamin D and normal or low-normal 25O-HD concentrations51.
The Committee concluded that a AI of vitamin D in the first year of life is 400 IU/day. This recommendation assumes that limited or no cutaneous synthesis of vitamin D occurs.
It is clear that during pregnancy and lactation, adaptive mechanisms come into play which reduce the mother's dependence on vitamin D for maternal calcium and bone homeostasis. During pregnancy the fetus is largely protected from the effects of maternal vitamin D deficiency, although there is evidence that fetal skeletal mineralization may be marginally affected and that neonatal hypocalcaemia is more common. There is also some evidence that fetal growth and infant birthweight might be adversely affected. Although there is circumstantial evidence that maternal vitamin D deficiency might influence bone development in later childhood, the evidence is insufficient to draw firm conclusions. Maternal vitamin D deficiency is, however, a major risk factor for vitamin D deficiency rickets in the breastfed infant, and thus efforts should be made to ensure maternal vitamin D adequacy during pregnancy. While the infant is breastfed, infant vitamin D status is effectively ensured through routine vitamin D supplementation (400 IU/day); however, if this method of supplementation is not favoured, maternal supplementation can also achieve desired infant concentrations, but larger doses to the mother are required and the long-term safety of these large doses has not been established.