The placenta and the fetal liver play a pivotal role in fetal Zn accretion, transfer, and utilization. The proximal bowel is responsible for postnatal accretion. The overall goal is to maintain homeostasis, and steady Zn tissue provision according to the needs of the developing fetus and newborn. The transfer of Zn at the cellular level and among tissues occurs via specific Zn-irk like receptors- ZIP, Solute Carrier family 39A (SLC39A) and receptors Zn transporters (ZnT) (SLC30A), the role of which will be expanded later in this review (Figure 1).

Measurement of the level of serum Zn, despite representing a subfraction (∼0.1%) of the total Zn pool, is considered the gold standard for monitoring its deficiency. The accretion of Zn occurs mainly during the second and third trimesters and hence extreme preterm newborns are at high risk of Zn deficiency, despite having high serum Zn levels at birth (53). The American Society for Parenteral and Enteral Nutrition (ASPEN) recommends a normal range for Zn concentration of 0.74–1.46 μg/ml (54). Normal serum Zn levels decrease with postnatal age (32, 55–57).

Several investigators including our group have shown that levels <0.7–0.74 μg/ml (deficiency) are associated with abnormal outcomes [poor growth, retinopathy of prematurity (ROP) etc.] (58, 59). The upper limit of normal (>1.46 μg/ml) has not been validated in large studies. UA and UV Zn levels have been utilized to estimate fetal Zn content and as explained earlier UV Zn levels are higher than UA levels, do not always correlate with maternal Zn levels and are affected by conditions associated with the birthing process (60). In a systematic review, cord Zn levels were found to be lower in pregnancies with various pathological outcomes [small for GA (SGA), IUGR, preeclampsia, smoking, etc.] and to correlate with maternal Zn levels (R = 0.4365) (61).

Zn concentration in hair could be valuable since it correlates with maternal Zn stores (62); however, obtaining enough sample size for Zn measurement may be a challenge in very preterm newborns. Zn supplementation in preterm newborns increases serum levels (total and skeletal subfraction) of alkaline phosphatase (ALP) (63, 64), a Zn-dependent enzyme. Bone alkaline phosphatase derives mainly from osteoblasts and plays a critical role in bone mineralization (65, 66). Zn deficiency can be suspected in infants with low serum concentration of ALP in the absence of rickets (67). Interestingly, Zn increased the activity and the half-life of ALP in vitro (68), without interfering with transcription of ALP protein. These findings could provide a link between Zn, ALP, growth and bone mineralization. It would be compelling to use ALP as an indirect, surrogate method to monitor Zn status since ALP levels are low in association with Zn deficiency and increase with Zn supplementation. However, this method is unlikely to be effective in preterm infants, since ALP may decrease with Zn deficiency and increase with cholestasis, vitamin D deficiency or fractures (63, 64). This may explain why studies in preterm infants have not shown a correlation between serum Zn levels and ALP (66, 69, 70). In addition, Zn deficiency can be encountered more frequently when DHM is provided to preterm neonates (40).

In summary: Measurement of serum Zn level (with normative values: 0.74–1.46 μg/ml) remains the most common and gold standard method for evaluation of Zn, despite its limitations. Serum Zn levels are affected by physiologic factors (GA, postmenstrual age, PMA, birthing process, growth etc.), but also from pathologic factors (inflammation, infection etc.). Umbilical serum Zn level should be used with caution (differences between UA, UV, not well standardized levels, influenced by birthing process etc.). Tissue Zn is ideal but difficult to obtain in neonates.

Zn is an important micronutrient that binds up to 10% of the proteins in the body and operates as a cofactor in a variety of enzymes (45). Zn interferes with or is structural part of enzymes and signaling pathways that as a net effect facilitate euglycemia, cellular cross talk, growth, brain functions and development. In these subsequent sections we summarize the evidence of specific enzymatic and biochemical pathways that could explain some of the benefits of Zn supplementation and some of the potential harms. The vast part of this evidence arises from preclinical animal studies or cell cultures.

Zn plays a critical role in the immune system. Acute dysregulation of Zn levels hinders the proliferation, maturation, and activation of cells in innate and adaptive immune responses, while chronically disrupted Zn homeostasis increases the risk of inflammation and disease (100, 101). Inflammation (fetal or neonatal) can affect fetal brain development as it relates with abnormal white matter growth and development of periventricular leukomalacia (PVL) (102, 103).

MTs can act as Zn acceptors and donors inside the cell, exchanging metal ions with proteins. MTs control the intracellular Zn levels with their sulfur cysteine groups, which release Zn after undergoing oxidation (123).

MTs provide age-dependent protection against neuronal toxicity with higher protective effect with advancing age as shown in rodents (124). Oxidizing conditions promote the release of Zn while reducing conditions restore Zn binding to MTs (125). By upregulating the expression of the metal regulatory transcription factor 1 (MTF-1), Zn increases the synthesis of MTs (126). MTs, in turn, act as metal scavengers and prevent the Fenton reaction and reactive oxygen species (ROS) production by binding active redox metals (123, 127). However, the age-induced increase in ROS finally compromises the Zn-binding capacity of MTs (128). Similarly, Zn deficiency increases ROS and inhibits MT activity, resulting in compromised mitochondrial function and cytochrome c oxidase activity (129).

MTs can store and release Zn depending on cellular needs contributing to its homeostasis. The stored Zn is important for the rapid growth of the brain that occurs during the latter part of gestation. The localization of MTs in the brain was first described by Suzuki et al. (130), which identified that the onset of the system starts at about 21–22 weeks GA. MTs I and II also play a role also in the distribution of Zn in this phase as they are found in specific glial populations, located in the periventricular zones. These cells migrate towards the cortex starting at 21 weeks of GA till 35 weeks GA, with possible completion and maturation of the system up to 10th postnatal month (130, 131). The role of MTs in these cases, with clear co-localization with glial proteins might relate with processes that involve these populations such as myelin production and neuronal migration.

In Summary: MTs play a role not only as ROS scavengers and prevention of toxicity in the fetal brain but also maintain homeostasis of Zn and help match the energy and Zn demand in the developing brain and migrating glial cells.

Zn binds to insulin and contributes to its biosynthesis, crystallization, and post translational maturation (132, 133). Zn is transferred inside the insulin secretory granules of β-cells via the ZnT8 transporter. Lack of ZnT8 activity prevents insulin crystallization and secretion and may be associated with type 1 and 2 diabetes mellitus (DM) (134, 135). As mentioned in the previous section, Zn decreases the expression of pro-inflammatory cytokines of IL-1β, TNF-α, and IL-6. In Zn deficiency, the long-term activity of these cytokines results in apoptosis of β-pancreatic cells and insulin resistance.

Apart from insulin control, Zn also participates in glucose metabolism. By activating GLUT4 in cell plasma membranes, it facilitates the uptake of glucose by insulin-dependent tissues (136, 137). In addition, Zn can act as an insulin-mimetic, inhibiting forkhead box transcription factors (FOXO) and regulating necessary gluconeogenic enzymes (138). Lastly, Zn inhibits glucagon secretion by inhibiting voltage-gated channels in pancreatic α-cells (139–141).

The role of Zn concentration on hyper- or hypoglycemia is not well established in preterm newborns. Limited evidence from our group showed that after increasing dose of Zn in PN to recommended dose in 23–28-week GA neonates the number of hyperglycemic episodes decreased, without a change in hypoglycemic events (142, 143).

Finally, association of hyperglycemia and ROP (144, 145) could be explained by abnormal modulation of HIF-1 and VGEF, two key factors for the development of ROP which also dysregulate after hyperglycemic conditions (146–148).

In summary: Zn affects glucose metabolism through insulin maturation, GLUT4 activation and glucagon inhibition. Zn sufficiency could improve glucose metabolism, decrease hyperglycemia, and improve downstream abnormal signals related to ROP. These effects need further exploration.

Superoxide dismutases (SODs) are crucial antioxidant enzymes that protect cells from ROS that arise in oxygen-rich environments from mitochondria, peroxisomes, and cytoplasm. SODs catalyze the conversion of superoxide to oxygen and hydrogen peroxide. Hydrogen peroxide is then eliminated by other antioxidant enzymes, such as catalase and glutathione peroxidases (149).

The catalytic properties of SODs depend on their metalation and disulfide bonding during posttranslational modifications (150). Zn and Cu are the catalytic metal ions for the human Cu-Zn-SOD (SOD1). The binding of the SOD1 homodimer to Zn offers structure stability, whereas Cu is responsible for enzymatic functionality (151). A stable connection with the His63 residue keeps Zn and Cu together and ensures the functionality of SOD1 even at extreme pH levels.

If SOD1 does not function normally, high ROS levels will cause oxidative damage, such as protein carbonylation, DNA breakage, and membrane lipid peroxidation (152). In adults, these can lead to different diseases, from cancer and amyotrophic lateral sclerosis to Parkinson's disease (153–155). Oxidative stress can also act as a stimulus for SOD1, so that it acts as a transcription factor. When hydrogen peroxide levels are high, SOD1 is phosphorylated via the cascade of Mec1, DNA Damage Response (DDR) kinase. The phosphorylated SOD1 then translocates to the nucleus, where, by binding to gene promoters, it controls the transcription of genes related to oxidative stress resistance (156).

SOD1 mutations can affect RNA metabolism. Mutant SOD1 can bind to the mRNAs of Neurofilament Light Chain (NFL) and Vascular Endothelial Growth Factor (VEGF) and form ribonucleoproteins complexes that then aggregate (157, 158). One of these RNA-binding proteins is HuR, which is protective against stress in motor neurons. Its interaction with the mutant SOD1 impairs its neuroprotective function and, along with the downregulation of VEGF, contributes to cytotoxicity and the appearance of neurodegenerative diseases like amyotrophic lateral sclerosis (159, 160).

In summary: Zn is the catalytic metal ion for SOD, a key antioxidant enzyme. If SOD is not adequate, high ROS levels can emerge resulting in oxidative stress, DNA breakage, lipid peroxidation and neuronal apoptosis.

Zn is abundant in various brain parts, especially the hippocampus, cerebral cortex, amygdala, olfactory bulb and retina (161). It plays crucial roles in synaptic transmission, neuronal signaling, and brain development in these regions. Zn deficiency or excess can result in abnormalities in neurodevelopment, behavior, CNS formation, and a variety of neurological diseases.

The hippocampus, which contains a high concentration of Zn, acts as the center of learning and memory and houses neuronal machinery responsible for stress responses. Accordingly, hippocampal Zn appears to be directly associated with learning, memory, and behavior (162, 163). Zn interacts with ligand-gated ion channels post-synaptically and modulates ion transport. Glutamatergic Zn-enriched neurons (ZENs) contain high densities of the excitatory glutamate receptors N-methyl-D-aspartate (NMDA) and alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), which have several modulatory Zn-binding sites (163). ZENs are found in the hippocampus, the olfactory bulb, and the dorsal cochlear nucleus (162) (Figure 2).

The NMDA and AMPA glutamate receptors regulate various calcium-, sodium-, and potassium-dependent intracellular pathways. In certain brain regions, namely the neocortex, the amygdala, and the hippocampus, their function is greatly affected by the extracellular Zn concentrations (164). Zn has a dual effect and can bind differently to the glutamate receptors, attenuating or amplifying their signals, depending on their levels. Because NMDA receptors are pivotal in triggering long-term changes in synaptic efficacy, it is evident that Zn participates as a co-transmitter in cognition and plasticity of the synapses (128). Apart from glutamate receptors, Zn also interacts with gamma (γ)-aminobutyric acid (GABA) receptors (165), glycine receptors and metabotrophic Zn receptors (166–170).

Preclinical data show that gestational Zn deficiency in rodents caused significant alternation in the production of and expression of NMDA subunits, brain derived neurotrophic factor (BDNF) and nerve growth factor (NGF) which caused memory and learning impairments later in life (171).

In summary: Zn affects synaptogenesis, neuronal plasticity and synaptic efficacy and in these ways plays a role in memory and behavior.

Neurogenesis, the CNS development process, occurs during gestational and neonatal periods. During this period, the radial and tangential migration and the division of neural stem cells (NSCs) create the neural tube, the neural crest, and the notochord. More specifically, the invagination of the neural plate creates the neural tube, and this process is called neurulation (172). When the neural tube closes, the neuroepithelial cells of the ventricular zone become radial glial progenitor cells (RGCs), which later turn into glial cells and neurons (173).

Zn, in the form of ZFPs, plays a crucial role in brain morphogenesis. ZFPs are involved in transcriptional regulation, signal transduction, actin targeting, and cell migration (96, 174). During neurogenesis, ZFPs direct the fate of embryonic stem cells (175). Zac1 protein participates in neuronal development in the cerebellum, Zn finger and BTB domain-containing protein 20 (Zbtb20) in the hippocampus, and fasciculation and elongation zeta protein (Fez) f1 and Fezf2 in the brain's olfactory region (176). ZFPs, e.g., Glioma-associated oncogene family Zn finger 3 (Gli3), regulate the cell cycle of RGCs, affecting their differentiation into neural cells. Interaction with the Sonic hedgehog protein (SHH) can alter the length of the G1 phase in the cell cycle of RGCs (153, 177). Mutations in Gli3 result in a shortened cell cycle, impaired cortical lamination, and cortical neuron formation and are responsible for various morphological CNS abnormalities, such as Greig cephalopolysyndactyly syndrome (GCPS), Pallister-Hall syndrome (PHS) and polydactyly and syndactyly (178).

Other characteristic examples are Zeb Zn fingers, Zeb1, and Zeb2. Zeb1 controls the development of the neocortex by acting as a transcriptional repressor that regulates the division mode of RGCs, their proliferation, migration, and differentiation. Zeb1 also affects the development of electrophysiological properties in developing neurons (179). Zeb2 participates in the differentiation of RGCs to Bergmann glial cells and astrocytes (180). Lastly, the Zic-type poly-ZNFs (Zic1, Zic2, Zic3, Zic4, and Zic5) take part in the closing of neural plates, the formation of the neural crests, the proliferation and the differentiation of NPCs in the medial forebrain, and cerebellar morphogenesis (181–183). Mutations in the Zic-type poly-ZNFs can hinder the physiological division of the brain into two hemispheres. They can also cause the Dandy-Walker malformation, which is associated with specific brain abnormalities such as cerebellar vermis hypoplasia and delayed motor and cognitive development (184).

In short, neurogenesis can be divided into three main stages: NSC proliferation and migration, neuronal differentiation, and cell survival (185–188).

Zn deficiency can affect multiple cell types in the brain.

Although Zn exists in many ocular tissues, such as the choroid, ciliary body, and iris, the highest Zn concentrations in the eye are found in the retina-choroid complex (464–472 μg/g Zn of dry weight) (222). Zn is an element necessary for many ocular metalloenzymes, both from a structural and functional aspect. Deficient Zn status can hinder ocular development, especially during the early prenatal stages (222). In the eye, vitamin A is converted to its active form, retinal, by a Zn metalloenzyme, alcohol dehydrogenase (ADH). Retinal is needed for the synthesis of rhodopsin, the photopigment found in the retinal rods responsible for night vision. Therefore, depressed ADH activity in cases of Zn deficiency would result in night blindness (222). In pregnant women with nyctalopia due to combined vitamin A and Zn deficiency responded to vitamin A and Zn supplementation but not to vitamin A or Zn alone (223). In addition, Zn has an important protective role against glutamate-induced toxicity in the retina. The retina is abundant in glutamatergic neurons. Zn is released along with glutamate in the retina, just like in the brain. By binding on the NMDA glutamate receptors of retinal ganglion cells and blocking excitation, Zn prevents the toxic effects of glutamate on the retina (224).

In preterm infants, the main concern is the possible association of Zn deficiency with the development of ROP. The pathophysiology of ROP evolves in critical hyperoxic and hypoxic phases with key enzymatic pathways such as VEGF, erythropoietin (EPO) and IGF-1 among others to play important roles in each phase (225–227). The role of Zn in the development of ROP is controversial and not well studied but could affect the development of ROP by several mechanisms: (1) as a potent antioxidant; (2) as modulator of glutamate receptors and amelioration of related toxicity as noted above, especially since glutamate can induce VEGF, a key molecule for development of ROP (228); (3) via modulation of transcription factors. Zn has antioxidant properties, as it is part of certain protective antioxidant enzymes, such as the previously mentioned enzyme SOD1, which has a protective effect against oxygen-induced retinopathy in mice (229). However, preterm neonates often have deficient Zn levels and subsequent reduced SOD activity. Moreover, Li et al. recently identified through a genome wide association study a novel lead single nucleotide polymorphism (SNP) that fell in an intronic region within the GLi3 gene. GLi3 is critical for the differentiation of the retinal pigment epithelium (RPE) and rod photoreceptor layer, functioning as both a transcriptional activator and repressor of canonical SHH signaling (230, 231). It also controls both the innate and adaptive immune response (232, 233), and, as aberrant inflammation is involved in the pathophysiology of ROP (234, 235), there is a possible association between GLi3 and ROP. The role of Zn related to EPO is quite complex, and to our knowledge indirect, and will be discussed briefly in the next paragraph. Clinically, exogenous human EPO initiated shortly after birth, although could result in angiogenesis, was not found to increase the risk of severe ROP (Preterm Erythropoietin Neuroprotection Trial, PENUT) (236).

Several clinical observational studies in premature newborns attempted to investigate the role of Zn with ROP, with so far mixed results. In one study, preterm newborns with ROP had lower maternal Zn levels, cord levels and serum levels at 40 weeks PMA when compared to those who did not develop ROP (237). In a retrospective study that involved infants 28–37 weeks GA, investigators measured one serum Zn level at less than 24 h after birth and in those newborns with levels <0.7 μg/ml, a higher proportion developed ROP when compared to those with higher Zn level (42% vs. 24%, P = 0.02) (59). However, in a systematic review of RCTs, enteral Zn supplementation compared to no supplementation or placebo had no effect on ROP (238, 239). Notably, several of the included studies also involved infants >32 weeks GA, with low baseline risk of ROP. The efficacy of maternal Zn supplementation for the prevention of ROP remains to be investigated.

In summary: Preclinical data link Zn deficiency with the development of ROP and other developmental eye disorders. Specifically, Zn could improve ROP via its antioxidant properties, modulation of glutamate pathway and excitotoxicity and via changes of downstream transcription factors. Nevertheless, robust clinical data to support this mechanistic link, is lacking so far.

Vitamin D is a lipid soluble vitamin with a steroid structure. Its structural integrity and functions are tightly regulated by Zn. Cholecalciferol is hydroxylated in the liver into 25-hydroxycholecalciferol, which is further hydroxylated in the kidney into 1,25-dihydroxycholecalciferol (active form). The latter interacts with vitamin D receptors (VDR). Once vitamin D interacts with its receptor, VDR dimerizes with the retinoid X receptor (RXR), to upregulate several downstream genes, including the calcium stimulated ATPase and alkaline phosphatase, causing increased intestinal absorption of calcium. VDR uses Zn to regulate the actions of vitamin D-dependent genes. In vitro, in the absence of Zn, VDR conformation cannot occur, and vitamin D function fails (240, 241). Zn supplementation increases Vit D levels and there a positive correlation between vitamin D levels and Zn levels (242, 243). Low Zn levels were found to predict vitamin D deficiency in adolescent women (244). On the other hand, vitamin D regulates Zn homeostasis by affecting the expression of Zn transporters (245). The role of vitamin D in normal brain function and development has been reviewed and demonstrated in several studies (246–248).

In summary: The intrinsic role of vitamin D in the regulation of Zn and vice versa should be taken into consideration when interpreting studies with Zn deficiency.

Zn participates in hemopoiesis in the form of ZFPs. Different ZFPs act on different cell lineages and maturation stages. ZFPs like GATA, Ikaros, FOG-1, Snail, MOZ, Gfi, and Zfp521 regulate the survival and differentiation of hematopoietic cells at the stem cell level until they commit to the lymphoid or myeloid lineage (249). Lymphopoiesis is regulated mainly by the ZFPs of the Ikaros family while myelopoiesis is controlled by a more complicated and diverse expression of ZFPs and has not yet been thoroughly studied. Among the 6 members of GATA-binding transcription factors of ZFPs, GATA-1 is important for erythroid cell line development (249). Zn is essential for lymphopoiesis and myelopoiesis, and it is also very important for erythropoiesis, mainly through the previously mentioned transcription factor GATA-1, the apoptotic proteins such as caspase 3, and the hypoxia-inducible factor (HIF) (250).

In the stage of proerythroblasts, GATA-1 release is triggered by the binding of erythropoietin to the erythropoietin receptor. Then GATA interacts with its co-factor, friend of GATA protein 1 (FOG-1), and together as a complex, they control gene expression (251). The formation of the complex requires the presence of Zn (252, 253). In addition, Zn, along with carnitine, inhibits the cleavage of caspase 3, a key component of apoptosis that is expressed when stimulation by erythropoietin is interrupted (254, 255). In this way, Zn prevents RBC apoptosis and promotes RBC survival, whereas Zn depletion results in RBC death (256, 257).

Erythropoiesis is initiated in hypoxic states when the HIF triggers the expression of the EPO gene. When oxygen levels are low, the HIF-1 subunit becomes stabilized and assembled. In contrast, in physiologic oxygen levels, HIF-1 is degraded. An HIF-1-specific prolyl-hydroxylase (PHD) hydroxylases proline-564 and/or −402 residues of HIF-1, signaling its ubiquitination and degradation by the proteasome. PHD2, also known as EGLN1, the main enzyme in control of hydroxylating HIF-1, has a conserved catalytic domain similar to that of other prolyl-4-hydroxylases. However, it also has a distinctive N-terminal MYND-type Zn finger domain (258) that has been anticipated to have either a positive or a negative regulatory function. Varying roles of this domain have been indicated, and according to the most recent results from Sinnema et al., the Zn finger ordinarily has a positive regulatory effect on the catalytic activity of PHD2. Specifically, a human PHD2 Zn finger mutation results in a loss of PHD2 function and is associated with congenital erythrocytosis (259).

HIF not only has functional roles but also takes part in the morphogenesis and development of the CNS. As shown by preclinical experiments, brain angiogenesis depends on the HIF signaling pathway since new vessel development is affected by oxygen regulation in neurons (260). In addition, HIF has a significant role in the neurogenesis of the autonomic/sympathetic nervous system (ANS), as the loss of HIF-1α hindered the survival and proliferation of pre- and post-ganglionic neurons. Specifically, lack of HIF-1α in the cardiac outflow tract, right ventricle and atrium, pharyngeal mesoderm, peripheral neurons, and hindlimbs caused hypoplasia of the sympathetic ganglion chain and diminished chromaffin cell number in the adrenal medulla (261). Astrocytes interact with HIF-1a in the opposite way; the deletion of HIF-1α prevents a hypoxia-induced cell death (262). Interestingly, the deletion of HIF-1α in neurons facilitates their hypoxia-induced cell death (262). Finally, HIF can induce hypomyelination by inhibiting the differentiation of oligodendrocyte precursor cells (263, 264).

In summary: The role of Zn in factors that affect hemopoiesis and lymphopoiesis is well supported by preclinical studies. HIF-1 is a regulator of erythropoiesis under conditions of hypoxia, but also a key regulator of brain morphogenesis, astrocyte function, myelination and vascular development of the CNS and ANS. Specific Zn containing enzymes (Zn-fingers) affect HIF-1 function, but whether Zn deficiency has significant effects on those pathways remain to be investigated.

Alcohol consumption may interfere with Zn tissue utilization and can have significant effects on the fetus (290, 291). Alcohol consumption can result in deficiency of several micronutrients including Zn. Whether Zn deficiency can contribute to the development of fetal alcohol syndrome (FAS) - a devastating disease with serious neurodevelopmental effects- is controversial (292). Zn is part of ADH, which detoxifies alcohol and generates aldehydes and ketones as metabolites. Alcohol in high concentrations can deactivate this enzyme (293). In addition, alcohol can induce increased urinary Zn excretion and low plasma levels in pregnant women (294), while alcohol abstinence reverses zincuria in adults (295). Affected infants from FAS could also present with low plasma and high urinary Zn excretion (296). In alcohol-exposed primates, when adequate Zn (3.5 mg/day) was given, maternal and neonatal brain Zn concentrations were similar to non-exposed controls, indicating that placental Zn transfer was not significantly affected (297). In rodents, whether alcohol affects Zn placental transfer is controversial (298, 299) and Zn supplementation might not be able to restore Zn transplacental movement in alcohol exposed mothers (300). In addition, ethanol-exposed rats had similar Zn placental transfer when compared to non-ethanol fed controls (301).

In summary: Alcohol consumption could be associated with multinutrient and/or isolated Zn deficiency. Although some of the manifestations of FAS may be related to co-existent or alcohol induced Zn deficiency, further research is needed to identify its significance in brain outcomes in the affected neonates.

The effects of smoking during pregnancy on fetal Zn are largely mediated by the Cd content of cigarettes. Cd and Zn are closely related in many metabolic pathways and often antagonize one another in their usage. Cd strongly induces MTs, which bind both Cd and Zn and can lower systemic Zn levels (302, 303). Smoking during pregnancy increases Cd levels in maternal blood and in placenta (302, 303). However, a large quantity of Cd is sequestered in the maternal liver and kidneys, with only a fraction accumulating in the placenta and the fetus (303). Mouse models have shown an increase in MT in the placenta after MT elevations in the liver have taken place, which tends to sequester Zn in the process, decreasing delivery of Zn to the fetus (52).

Additionally, cord blood analyses have shown that in mothers that do not smoke, elevations in placental Zn are associated with elevations in cord Zn, whereas, in mothers who smoke, elevations in placental Zn are not accompanied by elevations in cord Zn (302). Another notable feature of Cd's effect on maternal-fetal Zn delivery is through Zn transporters. In mouse models, Cd downregulates Zn transporters, which possibly could decrease Zn fetal uptake (303, 304). Mouse models have further associated maternal Cd exposure with a decreased BDNF and Zn level in the brain (304), indicating that smoking during pregnancy may impact fetal brain development. Supplementation of Zn appears to restore BDNF levels in fetuses (304).

In summary: Smoking affects Zn levels in the fetus primarily via Cd-mediated upregulation of MTs, and downregulation of Zn transporters, which together, act to sequester Zn in placental tissue and can harm fetal brain development as a downstream effect.

The neurodevelopmental effects of Zn in preterm newborns can be related to (1) its deficiency (low Zn level) (2) to its adequate supplementation or (3) to indirect effects on growth. Systematic reviews have shown that Zn supplementation increases growth and decreases mortality in preterm infants (238, 239). These findings are summarized in a Cochrane review (238) including one study with high-dose Zn intake [(305), ∼10 mg/day] and 4 studies with low (currently recommended) dose were included (63, 64, 306, 307). Other studies with high-dose Zn intake [(308, 309, 310) were not included]. The included studies were relatively small but had good methodological quality. The meta-analysis showed that enteral Zn supplementation may decrease all-cause mortality [Relative Risk (RR) 0.55, 95% Confidence Intervals (CI) 0.31–0.97; 3 studies, 345 infants; low-certainty evidence based on high-dose Zn study by Terrin et al. (305)], while had little or no effect on common morbidities such as bronchopulmonary dysplasia, ROP, bacterial sepsis, or NEC (low certainty of evidence). The authors concluded that Zn supplementation probably improves weight gain [standardized mean difference (SMD) 0.46, 95% CI 0.28–0.64; 5 studies, 481 infants; moderate-certainty evidence]; and may slightly improve linear growth (SMD 0.75, 95% CI 0.36–1.14, 3 studies, 289 infants; low-certainty evidence), but had minimal effect on fronto-occipital head circumference (FOC) (SMD 0.21, 95% CI −0.02 to 0.44, 3 studies, 289 infants; moderate-certainty evidence). No data was available on long-term neurodevelopmental outcomes at 18–24 months of age. In a systematic review that focused on anthropometrics and neurodevelopment only, Alshaikh et al. (239), included 8 RCTs studies (63, 64, 305–308, 311, 312) two of which utilized high dose (∼10 mg/day) Zn (305, 308). 7 out of 8 RCTs (742 infants), reported growth data at 3–6 months corrected age and 2 reported neurodevelopmental outcomes at 6–12 months, and will be further discussed later in this chapter (306, 312). Zn supplementation was associated with increased weight z-score (SMD 0.50; 95% CI 0.23–0.76), length z-score (SMD 1.12; 95% CI 0.63–1.61) and motor developmental score (SMD 9.54; 95% CI 6.6–12.4). There was no effect of Zn supplementation on FOC.

Most of the studies reported in both systematic reviews included older premature and low birth infants and as result, outcomes that occur more frequently in very low birth weight (VLBW) and extremely low birth weight (ELBW) infants and very preterm (28–32 weeks or extremely preterm (<28 weeks) infants (such as ROP, NEC etc.) were not assessed adequately.

Our group showed that In Zn-deficient ELBW preterm newborns, low dose Zn provision for at least 2 weeks improved the Z-score of FOC, but not length and weight (58). In another study in Japan, routine enteral Zn supplementation at 3 mg/kg/day, starting at 2 weeks of life to discharge without any PN Zn intake, showed no effect on growth parameters (313). In contrast, higher and incremental enteral doses of Zn (5–10 mg/day) in VLBW infants, improved linear growth in a study from India (314).

Differences in Zn dosing, duration of treatment and initiation and differences in baseline maternal Zn deficiency could all contribute to the variation in the above outcomes. Overall, though, Zn provision in premature infants has been shown to improve growth quite consistently (238, 239).

Isolated or balanced postnatal growth patterns in various parameters [FOC, length, weight, growth velocity and/or body mass index (BMI)] are all associated, and possibly independently, with improved neurodevelopmental outcomes in preterm infants (315–318). From those growth parameters, isolated linear growth in VLBW infants is associated with improved language scores in the Bayley-III scores and less chance for future developmental deficits (319, 320). Improved anti-inflammatory milieu and glucose-regulatory hormones (321–323) are important factors for adequate growth in preterm newborns. Zn provision could relate via its well-described anti-inflammatory and glucose regulatory properties towards this effect (148).

In summary: Several studies have shown consistently improved weight gain and linear growth in preterm newborns even after low or moderate (as currently recommended) Zn supplementation dosing schedules. Since normal growth can be independently associated with improved neurodevelopment, and Zn improves growth, it needs to be further investigated if and to what extent Zn supplementation can contribute or mediate a potential independent positive effect towards neurodevelopment.

So far, the direct link between Zn provision and neurodevelopment in newborns is not well established, and most data are derived from older infants and children. The overall effect of Zn supplementation in infants or children on neurodevelopmental outcomes is not convincing, possibly since brain development and growth is less dependent on Zn later in life.

Clinical data in infants and children: In infants, 1–12 months of age, Zn intake improves length and weight (324, 325) while higher Zn dose (10 mg/day vs. 5 mg/day) had a higher impact on those growth parameters (325). The effect of Zn supplementation—especially when administered alone- was investigated in a systematic review of 96 studies in 219,584 children (age: 6 months to 6 years) and showed benefits in Zn status, diarrhea- related morbidity, linear growth and possibly a small positive impact on all-cause mortality (289). Regarding neurodevelopment, Zn supplementation was found to improve minimally the executive function and motor development (326) or have no global effect (327), while supplementing mothers during pregnancy was not found to improve long term developmental outcomes (324, 325). On the other hand, Zn supplementation in infants who reside in areas with high baseline Zn deficiency, improved cognitive and sensorimotor developmental outcomes (328). A systematic review that included children 0–5 years (25 studies and 11,559 patients), showed no significant efficacy of Zn with and without Fe co-supplementation on child mental and motor development up to 9 years old age (329). Finally, a systematic review in children with attention-deficit/hyperactivity disorder (ADHD) has shown that Zn supplementation may improve total ADHD scores (330).

In summary: These studies in older infants and children show that the effects of Zn supplementation on neurodevelopmental outcomes are minimal and possibly more pronounced in individuals with baseline Zn deficiency.

In newborns, only a few RCTs have reported limited neurodevelopmental outcomes (Table 1). Of them, three reported neurodevelopmental outcomes at 6–12 months prematurity-corrected age and one at ≤3 months.

Ragab et al. (312) in a study completed in Egypt, used the parents-completed Ages and Stages Questionnaires (ASQ) at 4 and 6 months of life and showed improvement in communication, gross and fine motor skills, problem solving and social interaction in infants who received Zn supplementation. Friel et al. (306), conducted a study in the US in 1993 and reported a significant increase in motor developmental score using Griffiths mental development scale using Amiel-Tison neurologic assessment at 3, 6, 9 and 12 months of life (331). Mathur et al. (64), conducted a study in India and found that Zn provision improved alertness and attention pattern at 40 weeks post menstrual age and decreased hyper-excitability at 3 months corrected GA. As mentioned earlier, two of these studies (64, 306) were evaluated in a systematic review by Alshaikh et al. who showed an improvement in motor developmental scores, but not in global development (239).

In summary: The above-mentioned studies overall show benefits in early neurodevelopmental outcomes after Zn provision, although the time frame of observation is up to 12 months after birth, which does not meet current standard for follow-up to 22–28 months corrected age or school age.

Whether Zn levels could be associated with poor developmental outcomes is also poorly investigated. Terrin et al. (37) showed that in preterm newborns 23–34 weeks GA, there was a significantly positive correlation between total composite motor score (with Bayley III scale) and serum Zn levels at 28 days of life (DOL) (R = 0.467, P < 0.05) (33758400). The same authors found that serum Zn levels at 28 DOL were dependent on energy (β −0.650; P < 0.001) and protein (β −0.669; P < 0.001) intake received through PN in the first week of life.

In summary: Currently there is limited evidence to suggest low Zn levels (early, or later in life) relate to poor neurodevelopment. The interaction of Zn levels, energy consumption and neurodevelopment need further research.

Recent nutritional recommendations favored an increase in daily Zn supplementation (enteral Zn 2–3 mg/kg/day, parenteral Zn 500 μg/kg/day) (1–3). The enteral recommendation might not match physiologic data of transplacental fetal Zn accretion and actual needs for Zn might not be met, especially given the limited amount of Zn absorbed in the premature gut (32, 37). In addition, current recommendations, do not suggest different dosing schedules in cases of maternal Zn deficiency as commonly encountered in mothers with high poverty index and in women living in or recently immigrated from countries with endemic Zn deficiency (17). High Zn dosing (10 mg/dose) has been implemented by some investigators (305, 308, 332), but no neurodevelopmental outcomes have been provided so far. These studies have not reported major toxicity related to Zn and some have reported improved early outcomes, e.g., mortality, sepsis, necrotizing enterocolitis, feeding tolerance and/or growth. In Table 2 we summarize possible gaps of knowledge and future research directions.

In summary: We believe that well organized RCTs that include early Zn provision, at different - possibly higher - dosing schedules and include adequate neurodevelopmental follow up, need to be undertaken as soon as possible, in low-, middle- and high-income countries.