The dietary management of calcium and phosphate in children with CKD stages 2-5 and on dialysis-clinical practice recommendation from the Pediatric Renal Nutrition Taskforce

Developing the PICO questions

Recommendations in CPRs are most useful when they provide specific actionable advice on choosing between alternative approaches in particular clinical situations [5]. We developed clinical questions to be addressed by each statement and framed them in a searchable format, with specification of the patient group (P) to whom the statement would apply; the intervention (I) being considered; the comparator (C) (which may be “no action” or an alternative intervention); and the outcomes (O) affected by the intervention [5]. Our PICO terms were as follows:

• Population: Children from birth to 18 years of age with CKD2-5D
• Intervention: Nutritional requirements for Ca and P in children at different stages of CKD
• Comparator: Nutritional requirements for Ca and P in age-matched healthy controls
• Outcomes: Growth, bone disease, fracture risk, Ca balance, bone mineralization on imaging or biopsies, development of hypo- or hypercalcemia, hypo- or hyper-phosphatemia or hyperparathyroidism, and development of vascular calcification

The choice of P binder treatment is not within the scope of this document. The management of hyperphosphatemia in childhood CKD is covered in the Kidney Disease Improving Global Outcomes (KDIGO) 2017 CKD-MBD update [6] and the UK based National Institute for Health and Care Excellence (NICE) guideline on hyperphosphatemia management (published in 2013 and updated in July 2017) [7, 8]. We have not discussed dietary sources of vitamin D as natural (non-fortified) foods provide only a negligible amount of vitamin D [9] and do not significantly alter the serum levels of 25OHD or Ca [10].

Literature search

Details on the literature search are described in Supplemental Table 1.

Framing advice

After critically reviewing the literature for each question, we derived CPRs and graded them as suggested by the American Academy of Pediatrics (Supplemental Table 2) [11]. The strength of a recommendation is graded as strong, moderate, weak, or discretionary (when no recommendation can be made). The quality of evidence is graded high (A), moderate (B), low (C), or very low (D). Grade X refers to exceptional situations where validating studies cannot be performed and benefit or harm clearly predominate: in that case, a moderate or a strong recommendation may be given.

Using the Delphi method, voting group members were sent an e-questionnaire to provide a level of agreement on a 5-point scale (strongly disagree, disagree, neither agree nor disagree, agree, strongly agree) and given the opportunity for re-wording of recommendations if appropriate. Participants for the Delphi survey were selected based on the following three criteria: (i) a registered doctor or dietitian currently practising in pediatric nephrology; (ii) at least 5 years' experience with children and young people with CKD2-5D; (iii) currently associated with a major pediatric renal centre. It was agreed a priori that at least a 70% level of consensus was required for each statement, failing which the recommendation statement would be adapted after discussion in the core group, and reviewed again by the voting panel until a consensus level of at least 70% was achieved.

Evidence and rationale

Evidence and rationale

An estimate of usual Ca and P intake can be made using a variety of tools (Supplemental Table 3), each with advantages and disadvantages. Many of these have been developed for research purposes or for particular population groups, and vary in their accuracy as well as respondent and interviewer burden [50, 51]. Either a 3-day prospective diet diary/food intake record (semi-quantitative or weighed) or a food frequency questionnaire (FFQ) can give a clinically useful estimate of intake [52]. The 3-day prospective diet diary/food intake record is considered “gold standard” against which other dietary assessment methods are compared, but can be time consuming to administer and assess. FFQs tend to overestimate daily dietary nutrient intake, although this depends on the number of food categories included in the tool [53–55]. Supplemental Table 3 lists the relative advantages and disadvantages of the different methods of dietary assessment.

Clinicians may prefer to rapidly estimate calcium intake from a retrospective diet history of a typical 24-h period. This is a detailed dietary assessment technique consisting of questions about food and drinks consumed at meals and snacks through a typical 24-h period. This usually requires about 20–30 minutes by a dietitian or suitably trained designate. The history is usually taken from morning to night, capturing information on usual meal pattern and types, frequency, and quantity of foods consumed. Any additional nutrient source (e.g., nutritional supplements) should be included. Tools that can assist in estimation of portion sizes include use of common household measures (e.g., teaspoons, tablespoons, slices, and handfuls), food portion booklets [56], and photographic representations of portion sizes. A more detailed account of the frequency of consumption of specific dietary sources of Ca or P (such as milk, cereals, and additive-rich foods) can be ascertained by direct questioning. An estimate of Ca or P intake can be made from the diet history by reference to guides of the Ca and P content of common food items (see Tables 2 and 4). If a more accurate assessment is required, the portion sizes for each dietary item can be converted into weights and used with country specific food composition databases (FCDB) or dietary analysis software, supplemented by other sources, such as information from manufacturers, if required. The frequency of assessing the dietary intake should be guided by the child’s age, CKD stage, relevant nutritional concerns, or the presence of abnormal serum biochemistry.

Importantly, the Ca intake from P binders can contribute significantly to dietary Ca intake and must be included when calculating total daily Ca intake in CKD patients. There is little data on the Ca absorption from binders, particularly the effects of age and residual kidney function. Ca carbonate contains 40% Ca with an absorption of 20–30% and Ca acetate contains 25% Ca with an absorption of 20% when taken with meals [57, 58].

• 3.

  What are the Ca and P requirements in an infant, child and adolescent?

  • 3.1

    Requirements in healthy children

We describe the Ca and P requirements for healthy children as background and justification for estimating the requirements for children with CKD2-5D; specific recommendations for healthy children are outside the scope of this document.

Evidence and rationale

Adequate dietary Ca intake during childhood and adolescence is essential for normal skeletal mineralization [59]. The Ca content of the skeleton increases from 25 g at birth to around 1000 g in an adult female and 1200 g in an adult male. Approximately 25% of the total skeletal mass is laid down during the 2-year interval of peak height velocity during adolescence [60]. The normal levels of serum Ca and P have been previously described [41, 61, 62]. Balance studies indicate that dietary Ca absorption is the main driver of net Ca retention in infants and adolescents [63, 64]. The highest Ca requirements are during periods of rapid growth, including in the first year of life and during puberty, dropping after puberty to normal adult requirements [65]. The higher Ca requirements in young children are associated with higher normal values for serum total Ca, particularly during infancy, and reach adult levels by 5 years of age.

Randomized controlled trials (RCTs) in healthy children have shown that children require a higher Ca intake than adults, presumably for bone mineral accrual and growth [60, 66–68]. A double-blind placebo controlled RCT has shown that in pre-pubescent girls Ca-enriched foods significantly increased bone mass accrual over 1 year, with a preferential effect in the appendicular skeleton [66]. A further RCT has shown that in 12-year-old girls an increase in milk intake led to an increase in bone mineral density (BMD) and bone mineral content as measured by dual-energy x-ray absorptiometry (DXA) scan [67]. A RCT in pubertal girls with up to 7 years of follow-up showed that Ca supplementation significantly influenced bone accretion, particularly during the pubertal growth spurt, and the Ca requirement for growth was associated with skeletal size [68]. In fact, bone mineral accrual continues up to the end of the second or early in the third decade of life, depending on the skeletal site, when peak bone mass is achieved [60].

There is limited data on the Ca intake and health outcomes in normal children or suitable biomarkers of bone mineralization that can be used to derive suggested dietary requirements. An estimate of requirements for healthy infants is based on the Ca and P from breast milk. For children over 1 year of age, most international bodies use the factorial method to assess Ca and P requirements. This method accounts for the total quantity of mineral needed for bone accretion, growth, and replacement of obligatory losses (from stool, urine, and skin), and is adjusted for percentage absorption.

National recommendations for Ca and P intake for normal children have been reported by many countries [69–74] (Table 6). The international publications referenced in Table 6 use the most reliable methods for assessing mineral status. A number of different terms have been used in international recommendations to describe nutrient adequacy; these include Population Reference Intake (PRI), Recommended Dietary Allowance (RDA), Adequate Intake (AI), and others. Since these recommendations for dietary adequacy have different definitions and have used different methods in their derivation, some of the resulting recommendations differ widely. For example, for a 6–11-month-old infant, the Nordic Council of Medicine gives a Recommended Intake for calcium of 540 mg, whereas for a 6–12-month old, the US Institute of Medicine gives an Adequate Intake of 260 mg. We have taken a pragmatic approach and quoted the range of the published values for our recommendations. We refer to this new reference using a novel term, Suggested Dietary Intake (SDI; Table 7). The lower and upper limits of the SDI lie within the international published values for Ca and P intake in healthy children (Table 6), which represent the required average amounts of a nutrient plus 2 standard deviations. Currently, studies of dietary adequacy are not directly comparable because different standards have been used, depending on the publication used as the benchmark. As the SDI encompasses the current published international recommendations, future reference to the SDI as the benchmark for nutritional adequacy will allow better comparison between studies.

• 3.2

  Requirements in children with CKD2-5D

  • 3.2.1

    We suggest that the diet of children with CKD2-5D should be regularly assessed for total Ca and P content. The contribution of P additives to total P intake cannot be quantified, but dietary sources of P additives should be identified where possible. Frequency of assessment is based on the child’s age, CKD stage, and trends in serum Ca, P, and PTH (ungraded).

  • 3.2.2

    We suggest that the total Ca intake from diet and medications, including P binders, should be within the SDI, and be no more than twice the SDI, unless in exceptional circumstances (grade C, weak recommendation).

  • 3.2.3

    In special circumstances, such as for infants with CKD or those with mineral depleted bone, a higher Ca intake may be considered with careful monitoring (grade C, weak recommendation).

  • 3.2.4

    We suggest that the dietary P intake of children with CKD should be within the SDI for age, without compromising adequate nutrition (grade C, weak recommendation).

Evidence and rationale

As with healthy children, children with CKD require adequate Ca for skeletal mineralization, particularly during periods of active growth. However, excess Ca is potentially detrimental and can lead to ectopic calcification.

Childhood CKD is associated with nearly universal disturbances in bone and mineral metabolism, [2, 6, 75–77] resulting in bone pain, deformities [75, 76], growth retardation [2, 76], and fractures [76, 78]. Abnormal bone micro-architecture and mineralization defects are common and strongly associate with Ca status: mineralization defects were present in 29% of children in CKD2, and nearly 80% in those with CKD4-5 [79]. Similarly, in a cohort of peritoneal dialysis (PD) patients over 90% had deficient bone mineralization [80]. Of 170 children and adolescents with CKD stages 2-5D, 6.5% sustained a fracture over a 1-year follow-up [81], and this associated with lower tibial cortical BMD, particularly in growing children [81], and low serum Ca levels [81]. A prospective cohort study reported that children with pre-dialysis CKD have a 2–3-fold higher fracture risk than their healthy peers [82]. P-binder treatment was associated with a decreased risk of incident fractures, independent of age, sex, eGFR, and PTH levels [82], but of note, 82% of the children who were on P binders received a Ca-based P binder, suggesting that the Ca intake from binders may have a protective effect against fractures.

Conversely, a high Ca intake may lead to vascular calcification [83–85], progressive vessel stiffness [85, 86], left ventricular failure [87], and sudden death [81]. Vascular calcification begins in pre-dialysis CKD [85, 86] and accelerates on dialysis [85, 88]. In young adults on dialysis, the coronary artery calcification score was shown to double within 20 months [83], and worsening of the calcification score was associated with a higher serum Ca-P product and prescription of Ca-containing phosphate binders [83, 84]. Thus, high Ca levels carry a significant risk of vascular calcification and treatment strategies must be carefully tailored to provide enough Ca for mineralization of the growing skeleton, but not too much to cause ectopic calcification and cardiovascular disease.

There are no studies to indicate the appropriate amount of Ca for a child with CKD, and it is very likely that this would need to be individualized depending on the patient’s age, growth, and rate of bone turnover. The main source of Ca in the diet is dairy, which is frequently restricted in CKD due to its high P content. Hence, dietary intake alone may not provide sufficient Ca. Furthermore, intestinal Ca absorption may be impaired due to dysregulation of the normal homeostatic mechanisms. Low levels of activated vitamin D may lead to reduced Ca absorption [9, 40], even in the presence of Ca deficiency [89]. Extrapolation of data from adult CKD studies is not appropriate in this field, as the growing skeleton of children has significantly higher Ca requirements than a mature, possibly osteoporotic, adult skeleton.

In the absence of evidence-based studies, it would be reasonable to provide CKD children with comparable amounts of Ca intake (including Ca from diet, P binders, and supplements) as their healthy peers. Ca intake from all sources should provide at least 100% of the SDI. There are no studies to set a safe upper limit for Ca intake for healthy children of different ages [69]; however, based on expert opinion, KDOQI suggested 200% of the dietary reference intake as a safe upper limit [61]. Thus, the maximum Ca intake we recommend is 2600 mg/day (i.e., 200% of the SDI for 11–17-year olds). This is marginally higher than the KDOQI recommended upper limit of 2500 mg Ca/day as the KDOQI use American DRI, whereas as our SDI is based on recommendations from current international authorities: EFSA, DACH, NCM, and IOM.

Research shows that P retention has a significant role in the development and progression of CKD-MBD, including P-associated vasculopathy [4, 83–85, 90] and defects in bone mineralization [80, 91]. In adults on dialysis, high serum P levels have been linked to vascular calcification and death [92]. Control of serum P is essential for attenuating the progression of secondary hyperparathyroidism and associated cardiovascular damage in children [83–85, 90, 93]. P retention begins early in the course of CKD, although phosphaturic hormones such as fibroblast growth factor 23 (FGF-23) may prevent high serum P levels in early CKD [94, 95]. Nevertheless, it is important to limit dietary P intake to within the SDI in the mild-moderate stages of CKD (except in children with renal tubular disorders) and to the lower limit of the SDI in patients with advanced CKD and persistent hyperphosphatemia or hyperparathyroidism. Whey-based infant formulas have a low P content and continued use of these in place of casein-based formula or cow’s milk beyond infancy may be appropriate. Renal specific formulas may be considered as they have a lower P content, but caution is advised as potassium and Ca intake will also be considerably reduced. Careful advice during transition onto solid foods can control the introduction of higher P foods, especially when the bulk of P is provided from an infant formula. Importantly, aggressive dietary P restriction is difficult since it has the potential to compromise adequate intake of other nutrients, especially protein and Ca; hence, P binders are often required.

• 4.

  Managing the Ca and P requirements in children with CKD2-5D

  • 4.1

    We suggest that intake of Ca and P is adjusted to maintain serum Ca and P levels within the age-appropriate normal range, without compromising nutrition. Changes in management should be based on trends of serial results rather than a single result, with integration of serum Ca, P, PTH, alkaline phosphatase, and 25-vitamin D levels (grade C, weak recommendation).

  • 4.2

    We suggest that children with CKD who have hyperphosphatemia or hyperparathyroidism will require further dietary restriction of P, potentially to the lower limit of the SDI, without compromising adequate nutrition. Advice to limit the P contribution from phosphate additives should be given. Use of P binders for further control of serum P and PTH levels is often required, in addition to dietary restriction (grade C, weak recommendation).

  • 4.3

    We suggest that children with persistent hypocalcaemia or a high PTH may require a Ca intake above 200% of the SDI for calcium for short periods and under close medical supervision. Calcium can be provided through Ca supplementation, together with vitamin D (usually both native and active forms), as well as other sources of Ca such as a high Ca dialysate (grade C, weak recommendation).

  • 4.4

    We suggest that children with persistent hypophosphatemia should have their dietary P intake increased. P supplements may be necessary in some patients, particularly those on intensified dialysis or with renal wasting of P (grade C, weak recommendation).

Evidence and rationale

The intake of Ca and P is titrated based on serum levels of Ca, P, PTH, alkaline phosphatase, and 25-vitamin D considered together, following trends in results rather than single values, and maintaining levels within the age-appropriate normal range. Normal Ca and P levels are age dependent and are higher in infants and younger children [96]. Given the complex inter-dependency of these CKD-MBD measures, it is important to consider trends in levels, particularly for PTH, rather than a single value, as suggested in the KDIGO CKD-MBD recommendations [6]. Ionized Ca is the ideal measure when assessing serum Ca. It is influenced by alterations in albumin, circulating levels of anions, and acid-base status that are common in dialysis patients [97], and thus may not correlate with total serum Ca. Albumin-corrected Ca is superior to uncorrected total Ca, and should be used to modify treatment if ionized Ca is not available.

The intake of Ca from all sources must be considered: diet, medications (including Ca-based P binders and Ca supplements) and dialysate. The Ca intake from the diet may not provide sufficient Ca, particularly as most Ca-rich foods also contain large amounts of P, and are likely to be restricted in CKD patients, as discussed in “Introduction.” Furthermore, as reviewed in “Introduction,” the bioavailability of Ca and P from different foods varies and also depends on the vitamin D status of the individual [9, 10, 40].

The Ca content of P binder medications is shown in Table 8. The indication, dosage, and titration of different P binders, as well as use of different dialysate Ca concentrations, is outside the scope of this guideline and discussed extensively by other guideline groups [6, 7, 61]. It is noteworthy that international guidelines in children recommend the use of Ca-based P-binders as first-line treatment for hyperphosphatemia to help meet their high Ca requirements [41, 57]. However, it is important to consider the wide variation in elemental Ca intake and bioavailability from different Ca salts [61]. In addition, when Ca-containing medications are given with food, the Ca absorption is significantly lower compared to when they are given between meals: Ca absorption from Ca acetate averaged 21 ± 1% when given with food vs. 40 ± 4% when the binder was given while the subject was fasting [58].

Hypocalcaemia, common in untreated CKD, may be seen in patients who present with advanced CKD, in infants who typically have a high Ca requirement, in patients on intensified hemodialysis, and patients receiving a calcimimetic. Treatment with a calcimimetic, similar to a parathyroidectomy, commonly causes “hungry bone syndrome” and often requires more than 200% of the SDI for Ca for prolonged periods to replenish bone Ca stores. A higher Ca intake can be provided through Ca supplementation, together with vitamin D (usually both native and active forms), as well as other sources of Ca such as a high Ca dialysate. Preliminary clinical evidence indicates that in the setting of the hungry bone syndrome exogenous Ca is directed to the bone and not to the vasculature; a high Ca intake is associated with increase in BMD and regression of vascular calcification [98, 99].

Limiting dietary P intake is challenging, and adherence to P binding medications can be poor, especially amongst older children and adolescents; an international registry has shown that 80% of adolescents on peritoneal dialysis have high serum P levels [75]. Limiting the dietary P intake to the lower limit of the SDI may be required if there is either persistent hyperphosphatemia or persistent hyperparathyroidism in the presence of normal Ca levels; a P restriction to the lower limit of the SDI corresponds to the KDOQI recommendation of 80% of DRI [61] for most age groups. It is likely that P binders will be required at this stage, as stringent dietary restriction may be unrealistic and can compromise the intake of protein and other nutrients. As described in section 1, it is important to consider the sources of P and aim to avoid processed foods that are likely to contain inorganic P additives (Table 5), which are almost completely absorbed [100]. Dietary supplements and over-the-counter or prescription medications are hidden sources of P, and may contain P salts in their list of inactive ingredients [101]. The use of P binders can allow a more varied diet and P intake. However, these should not be used as a fixed dose, but should be adjusted to reflect the P content of a meal or snack [102]. For optimal control of P absorption, regular careful review of the use and administration of P binders is essential. Children and their caregivers benefit from regular diet education sessions [103, 104], with help from renal dietitians, as well as doctors and nurses, to understand the complexities of their renal diet [102, 105, 106].

Severe and prolonged dietary P restriction can lead to hypophosphatemic rickets in children. In addition, children on intensified dialysis regimens, including frequent daily or nocturnal dialysis, can have excessive dialysate P losses [107]. While high P losses may be compensated by a higher P diet, some may require P supplementation.

• 5.

  Management of the CKD patient with hypercalcemia

  • 5.1

    Acute, severe hypercalcemia can be life-threatening and requires rapid medical intervention (grade X, strong).

  • 5.2

    In a child with persistent mild to moderate hypercalcemia, we suggest a stepwise approach with reducing or stopping Ca supplements, Ca-based P-binders, and native and active vitamin D and using lower calcium dialysate. Transient reduction of dietary Ca, without compromising adequate nutrition, may be necessary. Regular reassessment is required, especially when Ca intake is reduced below the SDI (grade C, weak recommendation).

Evidence and rationale

Acute, severe hypercalcemia can be life-threatening and requires acute medical management; this is beyond the scope of this guideline.

Mild to moderate hypercalcemia can often be effectively addressed by reducing or eliminating intake of Ca supplements, native or active vitamin D, and replacing Ca-based P-binders with Ca-free binders. In addition, the Ca intake from dialysate as well as the Ca intake from other medications (such as potassium binders) must be reviewed and adjusted as best as possible. An investigation into other causes of hypercalcemia may be warranted [108]. A persistently high PTH in the presence of high serum Ca suggests hyperparathyroidism, and a calcimimetic, such as cinacalcet, may be indicated [109–111]. Rarely, high calcium levels in the presence of normal PTH may be seen, such as in immobilized children; in this situation, short-term bisphosphonate therapy may be considered to transiently lower serum Ca. Hypervitaminosis A is also known to cause hypercalcemia, particularly in children on dialysis [112], and must be considered in children with refractory hypercalcaemia. Frequent monitoring of serum Ca, ionized Ca, PTH, and 25-vitamin D levels, and following the trends in levels rather than a single value, is important.

If hypercalcemia persists after reviewing and adjusting medications and dialysate Ca (if appropriate), there is also a need for reduction of dietary sources of Ca and vitamin D, without compromising nutrition. A detailed dietary assessment to determine the key sources of Ca is required (see Table 2). As dairy products usually provide the largest source of Ca in a child’s diet, controlling portion sizes, or finding suitable replacements for cow’s milk, yogurt, and cheese is a primary goal for reducing dietary Ca intake. For infants and young children, a complete low Ca specialized formula (or a formula blend) may be necessary if hypercalcemia is severe and persists. These formulas can be used as a supplementary drink for older children. Plant-derived “milk” products made from oats, soya, almond, coconut, hemp, or similar products that have not been Ca-fortified can be useful for older children, but are often lower in energy and protein. It is advisable not to give rice milk to infants and young children due to its high arsenic content. Limiting intake of other Ca-rich foods may be necessary, including some fish, green-leafy vegetables, beans, nuts, and Ca-fortified foods and drinks. The replacement of tap water with deionized or distilled water also reduces Ca intake. Advice should be individualized, and seeks to minimize compromising the intake of other macro- and micronutrients.

Of note, fortification of foods with vitamin D, in particular dairy products, margarine, and some breakfast cereals, is increasingly common in several countries, and may inadvertently lead to a large cumulative intake of vitamin D, with ensuing hypercalcemia [10]. A careful assessment of food labels to determine if foods are vitamin D fortified is important and parents can be instructed to perform this. The Ca intake from all other sources, including all medications and dialysate, must be carefully recorded and adjusted if appropriate; the reader is referred to other international guidelines for further details [6, 8, 61].

Summary of recommendations

A summary of recommendations is provided in Table 9.

Research recommendations

We recommend the following areas of study to provide future evidence-based recommendations for Ca and P requirements in children with CKD2-5D:

1. Ca balance studies in children 1–18 years with CKD2-5D to determine the dietary Ca requirements for normal bone mineralization.

• 2.

  Ca balance studies to determine the absorption of Ca from Ca carbonate and Ca acetate in children with CKD2-5D who are vitamin D replete.

• 3.

  The effectiveness of the 24-h dietary recall as a tool to assess the Ca and phosphate intake in children compared with a 3-day diet diary (semi-quantitative or weighed) or a food frequency questionnaire.