Vitamin D and calcium intakes in general pediatric populations: A French expert consensus paper

Objectives: Nutritional vitamin D supplements are often used in general pediatrics. Here, the aim is to address vitamin D supplementation and calcium nutritional intakes in newborns, infants, children, and adolescents to prevent vitamin D de ﬁ ciency and rickets in general populations. Study design: We formulated clinical questions relating to the following categories: the Patient (or Population) to whom the recommendation will apply; the Intervention being considered; the Comparison (which may be “ no action, ” placebo, or an alternative intervention); and the Outcomes affected by the intervention (PICO). These PICO elements were arranged into the questions to be addressed in the literature searches. Each PICO question then formed the basis for a statement. The population covered consisted of children aged between 0 and 18 years and premature babies hospitalized in neonatology. Two groups were assembled: a core working group and a voting panel from different scienti ﬁ c pediatric committees from the French Society of Pediatrics and national scienti ﬁ c societies. Results: We present here 35 clinical practice points (CPPs) for the use of native vitamin D therapy (ergocalciferol, vitamin D 2 and cholecalciferol, vitamin D 3 ) and calcium nutritional intakes in general pediatric populations. Conclusion: This consensus document was developed to provide guidance to health care professionals on the use of nutritional vitamin D and dietary modalities to achieve the recommended calcium intakes in general pediatric populations. These CPPs will be revised periodically. Research recommendations to study key vitamin D outcome measures in children are also suggested.


Introduction
Vitamin D (vitD) is one of the hormones along with parathyroid hormone (PTH) and fibroblast growth factor 23 (FGF23) that upregulate circulating levels of phosphate and calcium (Fig. 1) [1].VitD deficiency is prevalent worldwide in general populations [2].
VitD and calcium intake prevent rickets in babies, children, and teenagers.Guidelines were published by the Nutrition Committee of the French Society of Pediatrics in 2012 [3] and by the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition in 2013 [4].
Based on new data constantly emerging in the field, the present manuscript addresses vitD supplementation and calcium nutritional intakes from birth to teenage years.We present 35 clinical practice points (CPPs) for the use of native vitD therapy in general pediatric populations (Table 1).

Methods
This consensus document was developed to provide guidance for health care professionals on the use of nutritional vitD in general pediatric populations (indications, dosing, and monitoring).These CPPs reflect the available evidence from clinical studies, expert opinion, and extrapolation from adult studies when appropriate.

The consensus development group
Two groups were assembled.The core group comprised pediatric nephrologists, endocrinologists, neonatologists, nutritionists, dieticians, research scientists, and a geneticist, all responsible for defining the scope of the project, formulating key questions, performing literature reviews, rating the quality of evidence, grading recommendations, and forming the voting panel.The voting panel included representatives from the French Society of Pediatrics, the French Societies of Neonatology, Pediatric Endocrinology, Pediatric Rheumatology, Pediatric Nephrology, Pediatric Gastroenterology, Hepatology and Nutrition, and the French Association of Primary Care Pediatricians.They also included representatives from the French Board of General Physicians with university positions and a biochemist specialized in vitD.Conflicts of interest are declared.

Developing the PICO questions
To give specific actionable advice, we developed clinical questions to be related to the following categories: the Population to whom the recommendation will apply, the Intervention being considered, the Comparison, and the Outcomes affected by the intervention.These PICO elements were arranged into questions to be addressed in the literature searches.
The population covered consisted of children aged 0−18 years in general pediatric populations and premature babies.The intervention was native vitD supplementation and the comparator was placebo or no supplementation, or intra-child comparison using the baseline status, before starting native vitD supplementation.The outcomes addressed were mostly changes in 25 hydroxy-vitD [25(OH)D] levels.Important safety outcomes, including changes in serum and urinary calcium levels, are presented when available.Most of the pediatric studies available did not investigate key patient outcomes such as bone fractures, cardiovascular events, or mortality, and we were unable to discuss these further but consider them as important topics for future research.The overall evidence is not very strong because of vitD supplementation has been used in pediatrics for decades, and because of the nonethical design of a clinical trial against placebo in the field.We addressed statements for indications, contraindications, treatment schedule, monitoring of native vitD supplementation, as well as specific "at-risk" situations in children, neonates, and premature babies.

Search strategy and selection criteria
The PubMed database was searched until October 2020: articles included were randomized clinical trials (RCTs), prospective uncontrolled or observational studies irrespective of the number of patients, registry data, retrospective studies, and reports with more than five pediatric patients, restricted to human studies in English.The literature search (PubMed) using vitD AND supplement filtered for the past 10 years and from birth to 18 years retrieved 1529 articles, 430 selected as reviews, case series, meta-analyses, or guidelines (Language: English), as summarized in Supplemental Table 1.If necessary, papers published before 2010 were also taken into consideration, but not exhaustively.

Grading system
We applied the grading system of the American Academy of Pediatrics (Fig. 2) [5].The quality of evidence is graded as high (A), moderate (B), low (C), or very low (D).Grading (X) refers to exceptional situations where validating studies cannot be conducted, and benefit or harm clearly predominates.
Voting group members were asked to [1] provide a level of agreement for the 35 CPPs included in this manuscript on a 5-point scale (strongly disagree, disagree, neither agree nor disagree, agree, strongly agree; Delphi method) using an e-questionnaire and [2] suggest rewording if appropriate.A consensus level of at least 70% was achieved for all CPPs.
Fig. 1.Overview of phosphate/calcium metabolism, adapted from [119] Green lines correspond to a stimulating effect Red lines correspond to an inhibitory effect FGF23: fibroblast growth factor 23; PTH: parathyroid hormone Briefly, vitD comes from diet and sun exposure, with a 25-hydroxylation in the liver and a 1-hydroxylation mainly in the kidney, but also in other target cells (see Fig. 3).VitD increases calcium and phosphate intestinal absorption, while inhibiting PTH and stimulating FGF23 synthesis.PTH is synthesized by parathyroid cells, mainly in response to changes in calcium concentrations through the regulation of the calcium-sensing receptor.PTH increases calcium levels, but inhibits tubular phosphate reabsorption, thus inducing hypophosphatemia.PTH stimulates both FGF23 and vitD synthesis.FGF23 is synthesized by osteocytes in the bone, and inhibits both PTH and 1,25(OH)2D while also inhibiting tubular phosphate reabsorption, thus inducing hypophosphatemia.FGF23 increases distal tubular calcium reabsorption, with a mild effect on calcium levels under physiological conditions.Klotho acts mainly as the co-factor of FGF23; it is synthesized by renal and parathyroid cells.[1].The activities of VitD 2 /D 3 are similar, but the pharmacokinetics and bioavailability of D 3 are greater [6].Apart from supplements, vitD content in the human diet is limited.Both 25(OH)D (halflife 2−3 weeks) and 1,25(OH) 2 D (half-life 4 h) levels result from a series of hydroxylations (Fig. 3) [7].Bioactive vitD activities are mediated through binding to the vitD receptor (VDR) (Fig. 4) [1].

VitD skeletal and extraskeletal effects
Rickets and hypocalcemia are caused by vitD deficiency, and by loss-of-function mutations in the VDR or in the vitD hydroxylases CYP27B1 and CYP2R1 [8,9].Conversely, hypercalcemia occurs because of vitD excess or CYP24A1 loss-of-function mutations [10].
The 25(OH)D levels provide information on the availability of bioactive precursors for target tissues, while 1,25(OH) 2 D levels reflect the activity of renal CYP27B1, the enzyme involved in the rate-limiting step of 1,25 (OH) 2 D synthesis.Both 25(OH)D and 1,25 (OH) 2 D are able to bind VDR and control biological processes, albeit with different affinities and potency [8].VitD also has anti-inflammatory and anti-proliferative activities [11].[7], measuring total 25(OH)D 2 / 3 , unless mentioned otherwise in reports.A few assays are not standardized, justifying always referring patients to the same laboratory [12].Some assays crossreact with the epimer-C3 forms, overestimating the amount of effective 25(OH)D, especially in preterm infants and neonates [13,14].A major staging effort was initiated during the last decade to move references from "normal values" to "recommended thresholds." The assay of 1,25(OH) 2 D is used to explore specific genetic, infectious, or inflammatory diseases [15].

Consensus statements
1. We recommend measuring only serum total 25(OH)D concentration in assessing vitD status in children.
2. We recommend measuring serum total 25(OH)D concentration in the same lab for a given child.
3. We recommend that assays of 1,25(OH) 2 D or other metabolites should not be used routinely in pediatrics.2) [16].This assay is not reimbursed in the absence of medical justification [17].If an underlying abnormality of phosphate/ calcium metabolism is suspected (hypo-or hypercalcemia), 25(OH)D levels should be measured.

Consensus statements
4. We do not recommend systematic measurement of serum total 25(OH)D concentration in general pediatric populations.
5. We recommend measurement of serum total 25(OH)D levels when there are symptoms of rickets.
Children with chronic diseases accumulate risk factors for both 25 (OH)D and calcium deficiency, which is why the target threshold of 25(OH)D level should be >30 ng/mL.Autopsy studies in adults showed that the threshold of 30 ng/mL was required to avoid mineralization defects [23].Two large epidemiological studies in adults associated increased mortality and 25(OH)D levels either >60 ng/mL or <20 ng/mL [24,25].In children, persistently elevated levels of 25 (OH)D likely increase the risk of hypercalciuria and subsequent nephrolithiasis and/or nephrocalcinosis [26,27].Thus, the American Endocrine Society and the Canadian Pediatric Society have proposed a target range of 30−60 ng/mL in children, to prevent both rickets and nephrolithiasis/nephrocalcinosis [16,28].

Consensus statements
6.We recommend a 25(OH)D level >20 ng/mL (50 nmol/L) in general pediatric populations to prevent rickets.
7. We suggest a 25(OH)D level >30 ng/mL (75 nmol/L) in general pediatric populations to avoid any mineralization defects and seasonal variability.

Rationale
Countries of the northern hemisphere recommend universal vitD supplementation in infants, toddlers, and adolescents to optimize 25 (OH)D levels and prevent nutritional rickets, despite the absence of large trials proving safety or efficacy for any disease outcome [3,4,21,29].

Effects in infants
VitD deficiency is common in mothers and infants, though the prevalence in diverse populations may depend upon sun exposure behaviors and supplementation during pregnancy [30].The persistence of maternal vitD deficiency/insufficiency during pregnancy/ breastfeeding irrespective of season and supplementation suggests that current vitD supplementation during pregnancy is inadequate [31].Human breast milk for healthy full-term newborns contains very little vitD even in vitD-repleted mothers [32].As such, exclusively breastfed infants, especially those born to vitD-deficient mothers, are at higher risk for rickets [33].Table 3 summarizes the RCTs of native vitD supplementation performed in infants [34−36].Briefly, the dose of 400 IU/day (1 mg = 40 IU) in healthy breastfed term babies prevents rickets but also ensures adequate bone health [37,38].Higher doses (1200 IU/day) are not associated with better outcomes [35].VitD supplementation ranging from 400 to 800 IU/day in healthy newborns is recommended in North America and Europe to maintain 25(OH)D levels of 10−20 ng/mL (25−50 nmol/L) [4].

Effects in healthy children and adolescents
Table 4 summarizes the RCTs on native vitD supplementation performed in healthy children and adolescents [39−49].A daily supplementation of 400 IU (vs.1000, 2000, and 4000 IU) is sufficient to prevent the 3-ng/mL physiological decrease in 25(OH)D concentrations over winter [39], so as to maintain 25(OH)D within the target Upon sun exposure, vitD 3 is formed by spontaneous isomerization of pre-vitD 3 , a precursor obtained by photo-isomerization of 7-dehydroxycholesterol in the skin.VitD enters the circulation after binding to the VitD binding protein (DBP), and therefore there is no specific storage in the liver [120].Of note, vitD 3 as well as its vegetal form vitD 2 can be obtained from diets enriched in fish and fungi, respectively, or from supplements.Once in the liver, vitD is hydroxylated at the carbon 25 by CYP2R1.Then a second hydroxylation step at carbon 1 by CYP27B1 in the kidney, the limiting reaction step, leads to the bioactive form of vitD, 1alpha-25-dihydroxy-vitD (1,25D).Importantly, the expression of these hydroxylases is controlled by the parathyroid hormone (PTH).Parathyroid glands (PTG) sense any variation of serum calcium levels, and in response to low levels, will produce and secrete PTH that in turn increases the expression of CYP2R1 and CYP27B1.In addition, 1,25D3 induces the expression of CYP24A1, a 24-hydroxylase, its main catabolic enzyme that converts it into calcitroic acid, an inactive hydrophilic degradation product eliminated by urine [121], thus completing the regulatory feedback loop of 1,25D 3 circulating levels.Of note, the metabolism of vitD in preterm infants and neonates is marked by an increase of a 25 OHD C3 epimer [122].In addition, 1,25D can be 3-epi hydroxylated.3epi-1,25D can bind the DBP with a modest affinity (36−46%) and the VDR with a low affinity (2−3%) compared to 1,25D.It is as effective as 1,25D in inhibiting PTH secretion or inducing phospholipid surfactant synthesis, but is less effective on CYP24A1 induction.Importantly, C3 epimer levels increase during the first 2 months of life and then decrease.Therefore, amounts of C3 epimers seem more relevant in preterm than in term newborns.However, the exact physiological role of these 3-epi vitD metabolites remains to be determined.range continuously.The mineralization peak is delayed as compared to the growth peak, thus providing a rationale to maintain supplementation throughout puberty [50].In healthy children, vitD repletion does not modify bone density, heart and vascular function, immune defense and infection, IGF1 levels, muscle strength, and physical performance [41,43,44,46−48,51−55].

Diet, daily intake, or single high-dose oral vitD (STOSS) therapy
In the absence of convincing evidence of efficacy and safety, daily vitD supplementation should be preferred in children, as proposed in other European countries [4], mainly to avoid increased urinary calcium following vitD administration, and to prevent nephrolithiasis.A recent RCT in children with chronic kidney disease and vitD deficiency showed that the time until 25(OH)D normalization and the number of children with 25(OH)D levels <30 ng/mL were similar after oral cholecalciferol for 3 months at various regimens, i.e., 3000 IU daily, 25,000 IU weekly, or 100,000 IU monthly [56].
The main risk for a daily regimen is likely noncompliance over the long term [57]; as such, given vitD pharmacokinetics, even though it has not been published yet in healthy children, one may consider weekly supplementation instead of daily supplementation, as proposed in particular populations [56].Still, because a significant number of children are at risk of rickets [58], administration of 50,000 −100,000 IU at regular intervals may also be considered in some children.However, the use of megadoses above 200,000 IU in a single shot should be avoided to prevent hypercalcemia/hypercalciuria and renal consequences [59−61].In any case, this rationale should be explained to parents, to provide them with the better choice for their child: intermittent supplementation will be preferred to no supplementation at all.

Novel concerns regarding vitD supplementation
A recent alert from the French Drug Agency (ANSES) reported the occurrence of severe hypercalcemia in neonates receiving food complements enriched in vitD instead of drug formulations [62], as previously reported [63].Particular attention should be given to food supplements containing vitD at concentrations sometimes seven to ten times that of licensed pharmaceutical formulations, significantly increasing the risk of overdose.Some parents' concern is the presence of excipients in the licensed formulation of vitD, leading parents and some physicians to propose food supplements instead of drug formulations.To date, there is no established link between the presence of certain flavors/excipients in the licensed preparations of vitD and the occurrence of diseases.The excipients vary according to the different forms, but each physician should have the free choice of  the prescribed specialty, if it is a licensed form of vitD, subject to manufacturing processes and controls performed by pharmaceutical companies.
VitD can be given any time during the day, and parents/patients should be warned that it may stick to glass or plastics.

Consensus statements
10.We recommend supplementing healthy children 0−18 years of age with a minimum of 400 IU of vitD per day.
11.We recommend supplementing healthy children 0−18 years of age with a maximum of 800 IU of vitD per day.
12. We recommend daily supplementation in children 0−2 years using D 2 or D 3.
13.We suggest preferring daily supplementation in children 2 −18 years using D 2 or D 3.
14.We suggest intermittent supplementation in the case of nonadherence in children 2−18 years using vitD 3 with either 50,000 IU quarterly or 80,000−100,000 IU twice in fall and winter.
15.We recommend avoiding of 200,000 IU of vitD in one shot.16.We recommend using only licensed pharmaceutical native vitD supplements.exposure, epidermal melanin composition, and obesity (Table 2).
Cutaneous vitD 3 synthesis depends on body location, extent of clothing, cultural factors, topical screen, latitude, altitude, season, pollution, and time of day for sun exposure [18,22].Levels of 25(OH)D and the increment in response to vitD supplementation are negatively influenced by adiposity [64,65].Table 5 summarizes the RCTs on vitD supplementation in obese children [66−70].
Because of current trends in dietary habits, the risk of vitD deficiency in individuals adopting a vegan diet should be emphasized.Since dietary vitD comes almost exclusively from fatty fish and fortified dairy products, subjects following a vegan diet excluding all animal products are at high risk of calcium/vitD deficiency and nutritional rickets [71,72].
Many chronic pediatric conditions involving intestinal malabsorption, inflammation, liver diseases, and kidney insufficiencies may also reduce vitD production and/or absorption.Drugs affect vitD metabolism through different mechanisms: enhanced catabolism (anticonvulsants), inhibition of intestinal calcium absorption (glucocorticoids), or induction of liver enzymes (nifedipine, spironolactone, clotrimazole, and rifampin) [73].To adjust vitD supplementation in these conditions, there are specific international guidelines [74−76].

Consensus statements
17.We recommend a minimum of 800 IU and a maximum of 1600 IU of vitD per day in children 2−18 years of age in the case of decreased availability of vitD (obesity, black ethnicity, absence of skin exposure to sun) or decreased intake (vegan diet).
18.In such children, we recommend daily supplementation with vitD 2 or D 3.
19.In such children, we suggest intermittent supplementation in the case of nonadherence using vitD 3 with either 50,000 IU every 6 weeks or 80,000−100,000 IU quarterly.20.We recommend considering at increased risk of developing rickets and vitD deficiency children and teenagers with the following conditions: malabsorption, maldigestion, chronic kidney disease, nephrotic syndrome, cholestasis, hepatic insufficiency, cystic fibrosis, secondary bone fragility, chronic inflammatory diseases, anorexia nervosa, skin diseases, anticonvulsant medications, or long-term corticosteroids.
21.We suggest that general pediatricians/physicians verify adherence to vitD supplementation in these children.
VitD intoxication is mainly caused by inappropriate prescription or administration of native vitD, and/or the use of high-dose overthe-counter unlicensed preparations, whether or not they are purchased on the internet [62,63].Circulating 25(OH)D levels are usually above 150 ng/mL (375 nmol/L) [16,63].

Consensus statements
22. We recommend that physicians rule out the use of over-thecounter vitD preparations before prescribing native vitD supplementation.
23.We suggest monitoring 25(OH)D levels in patients receiving treatment doses above the upper ranges currently recommended.
24.We recommend measuring 25(OH)D levels in the following conditions to adjust vitD supplementation: family history of vitD intoxication, hypercalcemia, hypercalciuria, kidney stones, and/or nephrocalcinosis.25.We suggest preferring daily vitD supplementation in these patients.

Calcium content and absorption efficiency of some foods
Calcium is absorbed in the intestine through nonsaturable passive paracellular absorption and by active transcellular absorption, respectively; absorption is greater in the duodenum and jejunum.In food, most of calcium is released from complexes with other dietary constituents to be absorbed [79].Increased growth velocity, dietary components, and a calcium-deficient diet improves the efficiency of calcium absorption.Calcium bioavailability results from the net calcium retention in bones and organs, the calcium absorption, and urine losses.
Dairy products are the major source of calcium due to an efficient absorption (30−40%), because of the lactose content [80].Fat (except in steatorrhea) and protein contents of the diet do not affect calcium absorption.
Many dietary components affect calcium absorption [79,81].Oxalate and phytate, through the formation of unabsorbable complexes, reduce the absorption of calcium when present in large amounts in some leafy green vegetables.Phytate is mainly present in whole wheat products.Conversely, oxalate content in brassica vegetables is very low, allowing efficient absorption of calcium.Insoluble fibers of fruits and vegetables (hemicellulose and lignin) impair calcium absorption because of their content of uronic acids [82].Some mineral waters are rich in calcium, their absorption efficiency being very close to that of dairy products (Table 6) [83].

Population reference intakes
The most recent European recommendations are from the EFSA (Table 7) [84].For infants 0−6 months, the EFSA published only mean calcium intakes based on breast milk: 200 mg/day [84].The previous 2011 American Dietary Reference Intakes (DRI) were higher; those of the French FSA in 2001 were intermediate.

Practical dietary modalities to evaluate and meet calcium requirements
Table 8 illustrates food equivalences necessary to ensure a calcium intake of 150 mg [83].The evaluation of calcium intakes should also consider the efficiency of calcium absorption (as compared to that of dairy products).Practical dietary modalities to meet calcium intake requirements according to age are illustrated in Table 9.Thus, from the age of 1 to 18 years, three or four dairy products are needed daily to meet calcium requirements.Consumption of mineral waters rich in calcium should be encouraged, especially when daily dairy product intake is insufficient.Replacing one or more dairy products with plants rich in calcium seems unreasonable since the amounts necessary to provide enough calcium would be too high, particularly at these ages or with plants rich in components affecting calcium absorption.
In exclusively breastfed or bottle-fed infants, calcium needs are covered.However, in case of vegan diets, only rice protein-based or soy protein-based infant formulas, whose composition complies with European regulations for infant formulas, cover calcium needs.All other plant-based beverages do not contain enough calcium to ensure adequate intake [71].The evaluation of calcium intakes should take into account the efficiency of calcium absorption (as compared to that of dairy products).For example, the efficiency of calcium absorption is seven-fold lower in spinach and rhubarb than in dairy products.The amount of calcium provided by these plants should therefore be divided by 7 when evaluating dietary calcium intakes.

Calcium deficiency
Calcium deficiency is defined by a prolonged calcium intake below the recommended intake for age.Its consequences are like those of vitD deficiency, namely insufficient intestinal calcium absorption to cover the needs of the organism despite reduced renal calcium excretion.However, dietary calcium deficiency is associated with secondary increased 1,25(OH) 2 D levels, which inhibits bone formation and redirects calcium towards the serum.This contributes to the maintenance of serum calcium when calcium transport is deficient, at the expense of bone mineralization [85−87].
A low calcium diet is associated with increased PTH levels in children with 25(OH)D levels >50 nmol/L [88].A chronic deficit in dietary calcium and increased PTH/1,25(OH) 2 D impair cartilage and bone mineralization [89].Furthermore, increased PTH levels inhibit tubular phosphate reabsorption, thus contributing to hypophosphatemia and further bone mineralization defect.

Criteria of calcium deficiency
Serum calcium levels are not a marker of calcium deficiency [58,87], but low urinary calcium (urinary Ca/creat <0.2 mmol/mmol) may be associated with low calcium intake [74].Assessment of daily dietary calcium intakes in toddlers, children, and adolescents leads to the three following categories [90]: sufficiency (>500 mg/day), insufficiency (300−500 mg/day), and deficiency (<300 mg/day).Such subjects with calcium deficiency should receive calcium supplementation from 250 to 1000 mg/day according to age, the total daily calcium dose not exceeding the population reference intake (PRI) for age.Diagnosis of calcium deficiency requires a dietary calcium intake evaluation, radiographs of wrists and knees, and measurement of plasma ALP, PTH, 25(OH)D, calcium and phosphate levels, and urinary calcium excretion.

Consensus statements
26.We recommend that, from the age of 1 to 18 years, children and adolescents should consume at least three to four portions of dairy products per day to cover calcium needs.
27.We recommend prescribing 500−1000 mg per day of calcium supplementation in children and adolescents receiving less than 300 mg adjusted for calcium bioavailability of nutritional calcium per day, especially in those following a vegan diet.
28.We recommend evaluating dietary calcium intakes in children with fractures and bone pain.
29. Diagnosis of calcium deficiency requires a dietary calcium intake evaluation, radiographs of wrists and knees, and measurement of plasma ALP, PTH, 25(OH)D, calcium and phosphate, and urinary excretion of calcium.

Evidence and rationale
Preterm infants are prone to vitD deficiency due to incomplete transplacental transfer during the third trimester, low body stores, low vitD in parenteral nutrition, decreased absorption, and negligible sun exposure during the hospital stay [91].At birth, 25(OH)D levels in infants are highly correlated with maternal levels [92].It is common for preterm infants to have low 25(OH)D levels [91,92]; respiratory distress syndrome [93], bronchopulmonary dysplasia [94], and enterocolitis [95] were associated with lower 25(OH)D levels at birth.Table 10 summarizes RCTs on native vitD supplementation in preterm infants [96−103].VitD intake through feeds varies greatly in preterm infants, from 2 (unfortified human milk) to 250 IU per day depending on the formula or the human milk fortifier chosen [104].Despite supplementation and nutritional support, rickets and metabolic bone disease are still observed in 8−50% of preterm infants at term corrected age (CA) [105−108].
Various results have been described in response to vitD supplementation of preterm babies: with 400 IU/day, 25(OH)D at term CA was <50 nmol/L in 65% of infants in India [90,99], while it was only 14% at 7 weeks CA in Ireland [109].With 800−1000 IU/day supplementation, 25(OH)D deficiency was less frequent but still present [96,103,110].
Beside preventing rickets, the challenge is also to avoid the onset of hypercalciuria and subsequent nephrocalcinosis, which is multifactorial in this particular population [111].About 8% of children who received 400 IU/day had 25(OH)D levels >125 nmol/L at 6 weeks CA [109].The majority of infants born before 28 weeks who received 1000 IU/day displayed 25(OH)D levels >150 nmol/L at 28 days of life [97].Lastly, despite serial measurements and therapeutic adjustment of doses, 25(OH)D levels were found to be >125 nmol/L in 19% of preterm infants at 34 weeks CA [112].A retrospective study found the threshold of 120 nmol/L for hypercalciuria and/or hypercalcemia [113], which was further confirmed prospectively (submitted data).Even though the level of evidence is low for optimal upper 25(OH)D levels in premature babies, and notably between 120 and 150 nmol/L, we suggest using 120 nmol/L in regards to the risk of hypercalciuria.
Data are scarce to establish post-discharge recommendations.Normal 25(OH)D levels have been characterized in 6-month-old infants who received supplementation of 1000 IU/day [114].One study suggested a lower incidence of wheezing before 12 months in babies who were randomly assigned to 400 IU/day of supplementation [115].For infants with abnormal 25(OH)D levels at NICU discharge, personalized supplementation may be discussed.10.2.Consensus statements 30.We recommend optimizing nutritional calcium and phosphate intakes in premature neonates.
31.We suggest that, during the initial stay in the NICU, preterm infants receive between 600 IU and 1000 IU per day of vitD, taking into account the content of vitD in milk and parenteral nutrition, vitD supplementation during pregnancy, and birth weight.
32.We recommend measurement of 25(OH)D in children born before 32 weeks of gestation or weighing less than 1500 g at 1 month of age.
33.We recommend 50 nmol/L as the lower target level and 120 nmol/L as the upper target level of 25(OH)D in premature neonates.34.After discharge from the NICU, we suggest following recommendations in general pediatric populations.

Evidence and rationale
There are no data on vitamin D status and supplementation of children from French overseas territories, whether from northern (e.g., Saint Pierre et Miquelon) or southern (e.g., Guiana, Antilles, Reunion Island, Tahiti, Mayotte) territories.Requirements for vitD supplementation notably depend on sunlight exposure, skin coloration, food contents, and individual genetic factors.Nutritional factors (such as calcium intake and phytate content), obesity, exclusive breastfeeding, and socioeconomic background may vary between Metropolitan France and the different overseas territories.
The current recommended dietary allowance for vitD (600 IU/day) may be inadequate in children residing in higher latitudes during winter to maintain 25(OH)D concentrations ≥20 ng/mL [40]: vitD intakes needed to maintain serum 25(OH)D concentrations at 12, 16, and 20 ng/mL in 90% of the children were 581, 1062, and 1543 IU/day, respectively.Systematic enrichment by vitamin D in food should also be taken into account, for example in Saint Pierre et Miquelon.
Skin pigmentation and genetic factors are also crucial for vitD metabolism, with a poor vitD status in many African-American children [116,117], with significantly decreased free and bioavailable 25

Research questions
Physiology and assays -What are the molecular and cellular mechanisms underlying 1,25D3-dependent calcium homeostasis?-Are extraskeletal effects observed in adults also applicable in children?-Is CYP27B1 regulated in the same way in children as in adults in all the target cells?-What other metabolites of vitD are interesting to evaluate?What are the target levels of 25-D levels?
-Are adult data adaptable to children?What should the supplementation schedule be in general pediatric populations?
-How should data be adapted from RCTs conducted in vitD-deficient subjects in general pediatric populations?-Could the weekly supplementation be used in general pediatric populations?-What are the real-life data (insurance database)?Which risk factors increase vitD dose?
-What would be the ideal vitD supplementation protocol in obese children, depending on their ethnicity?-How should data be adapted from RCTs conducted in vitD-deficient subjects in general pediatric populations?Which risk factors decrease vitD dose?
-What is the optimal schedule of vitD supplementation and monitoring in children and teenagers with hypercalciuria and nephrolithiasis?Nutritional calcium -How should the variability of calcium absorption be evaluated in children?-Is there a reliable laboratory parameter to assess calcium status and to recommend calcium supplementation in the case of deficiency?VitD supplementation in premature babies -What is the role of C3 epimerization in neonates and pregnant women?-What is the optimal schedule of vitD supplementation and monitoring in premature babies?-What is the ideal upper target of 25(OH)D levels in premature babies for bone outcomes but also global outcomes?-What is the exact frequency of vitD deficiency and overdose in very preterm infants?Overseas territories -Establish data on vitD status and needs for supplementation in overseas territories VitD: vitamin D.
(OH)D [118].In southern areas, two situations carry a risk of 25(OH)D deficiency: either children with dark skin or Caucasian/Asian children using sunscreen to protect their skin against the risk of melanoma and skin cancer.

Consensus statements
35.We suggest the same pattern of supplementation as in Metropolitan France.

Conclusion
We propose 35 CPPs on vitD/calcium supplementation in general pediatric populations.The overall policy of vitD supplementation must remain the prevention of nutritional rickets, but the "neither too much nor too little" rule should also avoid renal toxicity in the long term.Like other national and European guidelines, we propose daily supplementation of vitD.However, in case of poor adherence, we also propose an alternative protocol with intermittent administrations.New behaviors favoring the use of "more natural" over-the-counter forms of native vitD increase the risk of intoxication and misuse.Nevertheless, to prevent rickets and optimize bone health and peak bone mass, it is critical to provide all children aged 0−18 years with native vitD and adequate nutritional calcium intake, and to promote physical activity, in the setting of a shared clinical decision with the child and his/her parents.In the long term, it remains to be demonstrated whether vitD supplementation in children also has beneficial extraskeletal effects.Lastly, suggestions for research in the field are proposed (Table 11).
Supplemental Table 1: Literature review

Funding support
None to disclose in association with this manuscript.

Data sharing statement
Data sets and systematic literature review analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.

Declaration of Competing Interest
Dr. Tounian reports funding from CNIEL, Danone, Mead Johnson, M enarini, Nestl e, NHS, Nutricia, Sodilac.The other authors have no relevant conflicts to disclose.

3 .
Physiology and assays: relevance for clinical practice 3.1.Rationale 3.1.1.VitD as fat-soluble secosteroid VitD 2 and D 3 derive from ergosterol in plants and fungi, and from cholesterol in animals, respectively.VitD produced endogenously through the skin via sunlight accounts for 50−90% of circulating 25 (OH)D

3. 1 . 3 .
Assays for 25(OH)D and 1,25(OH) 2 D Circulating levels of 25(OH)D result from recent vitD intake and endogenous production.The gold standard technique for measurement of 25(OH)D uses the HPLC-tandem mass spectrometry (HPLC-MS/MS) technology.Automated techniques are now available at low cost

4 .
Which children should undergo 25(OH)D level measurement?4.1.Rationale Total 25(OH)D levels are usually measured in children with chronic diseases or with a combination of risk factors for vitD deficiency (Table

Fig. 2 .
Fig.2.Grading system for recommendations according to the American Academy of Pediatrics, adapted from[5].

Fig. 3 .
Fig. 3. Circulating 1,25(OH) 2 D levels are controlled by a tightly regulated networkUpon sun exposure, vitD 3 is formed by spontaneous isomerization of pre-vitD 3 , a precursor obtained by photo-isomerization of 7-dehydroxycholesterol in the skin.VitD enters the circulation after binding to the VitD binding protein (DBP), and therefore there is no specific storage in the liver[120].Of note, vitD 3 as well as its vegetal form vitD 2 can be obtained from diets enriched in fish and fungi, respectively, or from supplements.Once in the liver, vitD is hydroxylated at the carbon 25 by CYP2R1.Then a second hydroxylation step at carbon 1 by CYP27B1 in the kidney, the limiting reaction step, leads to the bioactive form of vitD, 1alpha-25-dihydroxy-vitD (1,25D).Importantly, the expression of these hydroxylases is controlled by the parathyroid hormone (PTH).Parathyroid glands (PTG) sense any variation of serum calcium levels, and in response to low levels, will produce and secrete PTH that in turn increases the expression of CYP2R1 and CYP27B1.In addition, 1,25D3 induces the expression of CYP24A1, a 24-hydroxylase, its main catabolic enzyme that converts it into calcitroic acid, an inactive hydrophilic degradation product eliminated by urine[121], thus completing the regulatory feedback loop of 1,25D 3 circulating levels.Of note, the metabolism of vitD in preterm infants and neonates is marked by an increase of a 25 OHD C3 epimer[122].In addition, 1,25D can be 3-epi hydroxylated.3epi-1,25D can bind the DBP with a modest affinity (36−46%) and the VDR with a low affinity (2−3%) compared to 1,25D.It is as effective as 1,25D in inhibiting PTH secretion or inducing phospholipid surfactant synthesis, but is less effective on CYP24A1 induction.Importantly, C3 epimer levels increase during the first 2 months of life and then decrease.Therefore, amounts of C3 epimers seem more relevant in preterm than in term newborns.However, the exact physiological role of these 3-epi vitD metabolites remains to be determined.

Fig. 4 .
Fig. 4. 1,25(OH) 2 D activities are mediated by the vitD receptorThe main physiological roles of bioactive vitD are to control calcium absorption in the intestine and calcium reabsorption in kidney, and under certain circumstances to promote calcium mobilization from bones[123].Such a tightly regulated network is principally controlled by the levels of vitD receptor (VDR) in the intestine, highlighting the importance of the 1,25(OH) 2 D/VDR signaling pathway in this tissue[87], In addition, several preclinical and clinical studies have highlighted the important extraskeletal role of 1,25(OH) 2 D and its potential therapeutic potency for the treatment of autoimmune disorders and various cancers.1,25(OH) 2 D activities are mediated through its binding to VDR (also known as NR1I1), a ligand-dependent transcription factor and member of the nuclear receptor superfamily.VDR is composed of a ligand-binding domain (LBD) and a DNA-binding domain (DBD) separated by a disorganized hinge region[10].Upon 1,25(OH) 2 D binding, VDR conformational changes allow VDR to translocate into the nucleus.Then VDR heterodimerizes with its partner retinoid X receptor to bind with high affinity to specific DNA sequences, called vitD response elements (VDRE), in the regulatory regions of 1,25(OH) 2 D/VDR target genes.After DNA binding, VDR conformational changes of helix 12 of the LBD favor the recruitment of coregulatory proteins to regulate the transcription of its target genes.Importantly, recent studies have demonstrated that in the absence of ligand, VDR is mainly cytosolic, but the residual VDR nuclear fraction is able to bind to VDRE and repress the expression of its target genes ((8,9), opening new avenues of the physiological consequences of low circulating 1,25(OH) 2 D levels.

7 .
RISK factors leading to increased native vitD supplementation 7.1.Rationale Many factors contribute to 25(OH)D deficiency, including increased bone turnover during periods of rapid skeletal growth, sun

Table 1
Summary and grading of the 35 consensus statements according to the American Academy of Pediatrics.
17We recommend a minimum of 800 IU and a maximum of 1600 IU of vitD per day in children 2 −18 years of age in the case of decreased availability of vitD (obesity, black ethnicity, absence of skin exposure to sun) or decreased intake (vegan diet)D 18In such children, we recommend daily supplementation with vitD 2 or D 3D  19In such children, we suggest intermittent supplementation in the case of nonadherence using vitD 3 with either 50,000 IU every 6 weeks or 80,000−100,000 IU quarterly D 20We recommend considering at increased risk of developing rickets and vitD deficiency children and teenagers with the following conditions: malabsorption, maldigestion, chronic kidney disease, nephrotic syndrome, cholestasis, hepatic insufficiency, cystic fibrosis, secondary bone fragility, chronic inflammatory diseases, anorexia nervosa, skin diseases, anticonvulsant medications,

Table 2
Risk factors of vitD deficiency in children.

Table 4
RCTs of vitD supplementation in healthy children and teenagers.

Table 3
RCTs of vitD supplementation in infants.

Table 5
RCTs of vitD supplementation in obese children.

Table 6
Calcium absorption efficiency depending on the type of food.

Table 7
EFSA: European Food Safety Authority.

Table 8
Food equivalents for calcium content.

Table 9
Practical dietary modalities to ensure whole daily recommended calcium intakes.

Table 11
Key research questions.