Serum copper concentrations in hospitalized newborns

Journal of Trace Elements in Medicine and Biology 39 (2017) 1–5

Contents lists available at ScienceDirect

Journal of Trace Elements in Medicine and Biology journal homepage: www.elsevier.com/locate/jtemb

Epidemiology

Serum copper concentrations in hospitalized newborns Ricardo González-Tarancón a , Luisa Calvo-Ruata a , Maite Aramendía b , Carmen Ortega a , Elena García-González a , Luis Rello a,∗ a b

Department of Clinical Biochemistry, Hospital Universitario “Miguel Servet”, Paseo Isabel La Católica 1-3, 50009, Zaragoza, Spain Centro Universitario de la Defensa-Academia General Militar de Zaragoza, Carretera de Huesca s/n, 50090, Zaragoza, Spain

a r t i c l e

i n f o

Article history: Received 11 January 2016 Received in revised form 27 June 2016 Accepted 30 June 2016 Keywords: Atomic absorption spectrometry Copper Ceruloplasmin High sensitivity C-reactive protein Newborn

a b s t r a c t Background: Low serum Cu and ceruloplasmin (Cp) concentrations in newborns can be the first indication of a severe Cu deficient intake or, alternatively, of genetic diseases affecting Cu metabolism. However, Cu and Cp concentrations can also be influenced by other variables that render their quantitative results difficult to interpret. Therefore, it is necessary to identify these variables and stratify Cu and Cp concentrations according to these altering factors. Methods: Serum Cu and Cp concentrations for 564 hospitalized newborns (0–12 days of life) are stratified according to their age, prematurity (birth weight or gestational age), type of feeding and inflammatory state (assessed by the serum high sensitivity C-reactive protein (hs-CRP) level) to identify potential correlations. Results: Serum Cu and Cp concentrations are influenced by all four variables analyzed, although inflammation is the most significant: the greater the hs-CRP concentration, the greater the serum Cu and Cp concentrations. Prematurity is also an important factor and preterm infants often show very low Cu and Cp concentrations. Age of life and type of feeding have in turn a more modest effect on these magnitudes, being slightly greater at 3–5 days of age in breastfed newborns. Conclusions: Inflammation and prematurity are the main variables affecting serum Cu and Cp concentrations in newborns. Therefore, hs-CRP should always be assayed in parallel to Cu status. When there is an inflammatory state proper interpretation of these concentrations can be challenging. © 2016 Elsevier GmbH. All rights reserved.

1. Introduction Copper is an essential trace element that is required for adequate human development [1]. In plasma, approximately 75–90% of Cu is bound to ceruloplasmin (Cp) [2], so Cu concentrations are parallel to those of Cp. Cu deficiency in newborns as a consequence of an inadequate dietary intake during pregnancy is rare. However, deficiency of this element can arise in the neonatal period by several means [1,3], the most important being produced by genetically determined conditions, such as Wilson’s Disease (WD) or Menkes Disease, or by a very low Cu intake in newborns fed with improperly Cu supplemented artificial formulas [4,5]. For this reason, paediatricians often request evaluation of Cu status in hospitalized newborns.

However, interpretation of serum Cu and Cp concentrations in this population is difficult because they also influenced by several variables, such as, prematurity [6,7], type of feeding [8,9] and inflammatory conditions [10]. This is the reason why the use of Cp to screen for WD in the neonatal period has proven unsuccessful [11] or why extreme Cu deficiencies in newborns fed with artificial formulas are often detected after elapsing several months [4,5]. In this paper the concentrations obtained for Cu and Cp in hospitalized newborns are presented. As shown below, proper interpretation of the quantitative results obtained requires the identification and stratification of the different variables influencing these concentrations.

2. Material and methods

Abbreviations: Cp, ceruloplasmin; EQAS, external quality assurance schemes; GFAAS, graphite furnace atomic absorption spectrometer/spectrometry; hs-CRP, high sensitivity C-reactive protein; LOD, limit of detection. ∗ Corresponding author. E-mail address: [email protected] (L. Rello). http://dx.doi.org/10.1016/j.jtemb.2016.06.010 0946-672X/© 2016 Elsevier GmbH. All rights reserved.

2.1. Patients 596 newborns were initially included in the study after obtaining written consent from their parents. All of them were born between January 2009 and December 2010 and were admitted at

2

R. González-Tarancón et al. / Journal of Trace Elements in Medicine and Biology 39 (2017) 1–5

the Neonatal Unit or the Neonatal Intensive Care Unit of the Hospital Universitario Miguel Servet (HUMS) for several reasons. The main causes of admittance were prematurity, neonatal jaundice, risk of infection, apnoea or respiratory distress, feeding problems or convulsions. In all cases, serum Cu and Cp concentrations were determined as part of a complete blood analysis study requested by the paediatricians in charge, which is included in the neurometabolic protocol established by the Paediatrics Department of the HUMS. Data from 32 babies who died in the following weeks/months after birth due to their severe pathologies and/or extreme prematurity were finally excluded from the study. According to their clinical histories, none of the 564 individuals finally included in the study developed signs of Cu deficiency after their stay at the Neonatal Units of the HUMS and, until November 2015, none has been diagnosed with WD. To observe the effect of the type of feeding on serum Cu and Cp concentrations, newborns were categorized in three groups: breast, mixed or artificial feeding. The artificial formulas provided contained approximately 3000 ± 20 ␮g Cu/L for term newborns and 5000 ± 20 ␮g Cu/L for preterm newborns. The content of Cu in human breast milk is approximately 1200–1500 ␮g/L [12], although in this case the range of milk Cu concentrations among mothers is very wide, and values ranging from 30 to 2190 ␮g/L have been reported [13]. In any case, the exact Cu intake in each newborn was unknown, as the amount of milk/artificial formula ingested was not documented. The HUMS Ethics Committee approved the study.

(OELM Serum, Trace Elements External Quality Assessment Scheme organized by the Société Franc¸aise de Biologie Clinique), ranged between −5.1% and +3.2%, with a total Z score mean of −0.4.

2.2. Instrumentation

The sociodemographic variables were categorized using established or study-specific cut-offs. Thus, prematurity was conventionally defined as birth weight or gestational age below 2500 g or 36 weeks, respectively. Days of life at sample collection were arbitrarily separated into 3 groups, 0–2, 3–5 and 6–12 days. To study the effect of the inflammatory state on serum Cu and Cp concentrations, serum hs-CRP concentration was also stratified defining 3 intervals: normal (<5 mg/L), slightly elevated (5–20 mg/L) and overtly high (>20 mg/L). All categorical variables (birth weight, gestational age and days of life at sample collection) are presented as medians (25th–75th percentiles). Normality of Cu and Cp concentrations in each subgroup was assessed using the Kolmogorov–Smirnov test. As these concentrations were not normally distributed in any of the cases, they are also reported as medians (2.5th–97.5th percentiles). Differences in Cu and Cp concentrations were assessed by UMann-Whitney or Kruskall-Wallis tests when comparing two or more than two subgroups, respectively. Differences with p values below 0.05 were considered to be statistically significant. Finally, correlation between serum Cu and hs-CRP concentrations was assessed using the Spearman’s correlation coefficient (in that case, serum hs-CRP concentration was considered to be a continuous variable). Statistical analyses were conducted with SPSS software (IBM Company, SPSS Statistics version 18.0, United States).

Cu was determined in a graphite furnace atomic absorption spectrometer (GFAAS) ZEEnit 600 with Zeeman correction (Analytik Jena, Jena, Germany); Cp concentrations were determined by nephelometry on an Immage 800 apparatus (Beckman Coulter, Brea, CA, USA) and hs-CRP concentrations were measured by turbidimetry on the AU analysers (Beckman Coulter). Serum was the specimen of choice in all cases. The analyses were made at the Department of Clinical Biochemistry of the HUMS. This laboratory participates in several national and international external quality assurance schemes (EQAS) and performs successfully in all three magnitudes analyzed. It is worth mentioning that the GFAAS method for Cu determination used in this work, adapted from the one published by Almeida and Lima [14], only requires a minimum quantity of serum (10 ␮L). When using Flame Atomic Absorption Spectrometry (FAAS), the analytical technique of choice in most laboratories for serum Cu measurement, hundreds of microliters of sample are required, a volume not always available for this population especially, as in this case, when multiple tests are requested. Serum samples were 100-fold diluted with a 0.028 M HNO3 solution and 10 ␮L of the diluted sample was directly introduced in the GFAAS spectrometer by means of an autosampler. Two replicate measurements were carried out for each determination. Optimized conditions for best detection power were achieved by deploying the Cu line at 324.8 nm, a pyrolysis T of 1200 ◦ C and an atomization T of 2000 ◦ C, which yielded a limit of detection (LOD) of 4.2 ␮g/L for the undiluted sample. This LOD is slightly poorer than those reported in other works measuring serum Cu by GFAAS: 4.0 ␮g/L in the work by Correia et al. [15] (1/80 sample dilution) or 0.98 ␮g/L in the work by Almeida and Lima [14] (1/25 sample dilution). This fact might be a direct consequence of the greater sample dilution factor deployed in this case (1/100). In any case, the LOD achieved is clearly fit for purpose, as all samples analyzed had Cu concentrations above the limit of quantification. Precision of the technique, evaluated from analysis of quality control materials, was 3.7% for 801 ␮g Cu/L and 6.2% for 1340 ␮g Cu/L. Bias, estimated by EQAS performance

2.3. Sampling conditions and contamination issues for Cu determination Following standard phlebotomies, venous blood samples were drawn into siliconized trace metal tubes (Vacutainer trace element serum, reference 368380. Becton Dickinson, Franklin Lakes, NJ, USA). After centrifugation, 100 ␮L serum aliquots were transferred to metal free polypropylene tubes (1.8 mL Nunc cryotube vials, catalog number 363401. Roskilde, Denmark) and immediately frozen at −20 ◦ C until Cu analysis. All Cu assays were performed less than 1 month after collection. All laboratory material used for analysis (glassware, pipette tips, autosampler cups) was previously immersed for 24 h in a 0.28 M HNO3 solution, followed by rinsing with Milli-Q water. As serum Cu concentrations in newborns are expected to be in the order of hundreds of ␮g/L, any additional precaution to avoid contamination was considered unnecessary. In fact, blank absorbance values were always in the range of 0.0020–0.0030 s−1 , while 1 ␮g/L Cu aqueous standard solutions provided absorbance values in the 0.0120–0.0140 s−1 range, a signal that can be easily differentiated from blank levels. 2.4. Statistical analyses

3. Results The main altering factor influencing Cu (and Cp) serum concentrations in newborns is inflammation, even more than prematurity. The characteristics of the group of patients under study according to the main variables considered are shown in Table 1. Tables 2 and 3 show the serum Cu and Cp concentrations obtained, respectively, for each of the newborn subgroups considered. In each of these subgroups, the greater the hs-CRP, the greater the Cu and Cp concentrations (pb value, last column in Tables 2 and 3). As expected, term babies (>2500 g or >36 weeks) have serum Cu and Cp concentrations greater than preterm newborns, but these

R. González-Tarancón et al. / Journal of Trace Elements in Medicine and Biology 39 (2017) 1–5

3

Table 1 Demographic variables for the newborns included in the study, expressed as median and 25th and 75th percentiles. Newborns were separated into three subgroups according to arbitrary ranges of hs-CRP concentration. Differences between subgroups are assessed by Kruskall-Wallis tests. hs-CRP (<5 mg/L) n Birth weight (g) ≤2500 130 >2500 188 Gestational weeks 152 ≤36 166 >36 Days of life 0–2 152 3–5 118 6–12 50 a

hs-CRP (5–20 mg/L)

hs-CRP (>20 mg/L)

Median

P25

P75

n

Median

P25

P75

n

Median

P25

P75

pa

1833 3170

1520 2875

2000 3500

62 84

1650 3125

1233 2850

1820 3780

32 52

1675 3300

1050 3010

1925 3620

0.139 0.081

33 39

31 37

34 40

70 74

33 40

30 39

34 40

38 46

32 40

31 39

36 40

0.614 0.112

1 4 9

0 3 8

2 4 10

50 66 36

1 4 8

1 3 7

2 4 9

38 26 26

1 4 10

0 3 8

2 4 11

0.268 0.548 0.058

p value (significant <0.05) for differences between hs-CRP subgroups within each row.

Table 2 Serum Cu concentrations (␮g/L), expressed as median and 2.5th and 97.5th percentiles. Differences between subgroups are assessed by U-Mann-Whitney or Kruskall-Wallis tests. Cu (␮g/L)

hs-CRP (<5 mg/L) n

Birth weight (g) ≤2500 130 >2500 188 pa 0.000* Gestational weeks 152 ≤36 >36 166 pa 0.000* Days of life 0–2 152 3–5 118 6–12 50 a p 0.000* a b *

hs-CRP (5–20 mg/L)

hs-CRP (>20 mg/L)

Median

P2.5

P97.5

n

Median

P2.5

P97.5

n

Median

P2.5

P97.5

pb

390 489

30 180

900 940

62 84 0.000*

530 774

199 490

2410 1470

32 52 0.068

696 875

416 400

2260 1700

0.000* 0.000*

400 500

30 180

900 940

70 74 0.000*

540 833

199 250

2410 2575

38 46 0.064

780 900

416 400

2260 1700

0.000* 0.000*

410 500 430

30 280 180

830 940 1060

50 66 36 0.109

580 680 833

240 450 199

2410 2575 1470

38 26 26 0.275

850 772 1057

400 416 476

2260 1700 1823

0.000* 0.000* 0.000*

p value for differences between subgroups for each hs-CRP group. p value for differences between hs-CRP subgroups within each row. significant p value <0.05.

Table 3 Serum ceruloplasmin (Cp) concentrations (mg/L), expressed as median and 2.5th and 97.5th percentiles. Differences between subgroups are assessed by U-Mann-Whitney or Kruskall-Wallis tests. Cp (mg/L)

hs-CRP (<5 mg/L) n

Birth weight (g) 130 ≤2500 >2500 188 pa 0.000* Gestational weeks 152 ≤36 >36 166 pa 0.000* Days of life 152 0–2 3–5 118 6–12 50 pa 0.000* a b *

hs-CRP (5–20 mg/L)

hs-CRP (>20 mg/L)

Median

P2.5

P97.5

n

Median

P2.5

P97.5

n

Median

P2.5

P97.5

pb

116 148

28 60

271 345

62 84 0.000*

161 258

62 72

555 453

32 52 0.052

226 282

156 150

438 417

0.000* 0.000*

117 157

28 61

239 345

70 74 0.000*

163 299

62 71

555 480

38 46 0.007*

218 283

156 150

438 417

0.000* 0.000*

121 148 122

28 80 60

303 345 390

50 66 36 0.008*

165 239 268

62 72 111

555 480 453

38 26 26 0.238

265 247 345

150 156 175

438 417 570

0.000* 0.000* 0.000*

p value for differences between subgroups for each hs-CRP group. p value for differences between hs-CRP subgroups within each row. significant p value <0.05.

differences are statistically significant only when the serum hsCRP concentration is within normal range or slightly elevated (pa value in Tables 2 and 3). When hs-CRP is >20 mg/L, the differences between term and preterm newborns are mostly attenuated. Fig. 1 shows the correlation between serum Cu and hs-CRP concentrations (hs-CRP log-transformed) for the subgroup of newborns >2500 g and 0–2 days of life. The Spearman’s correlation coefficient obtained for this population subgroup is 0.41.

The effect of the type of feeding on serum Cu and Cp concentrations was evaluated only for term newborns (>2500 g) whose hs-CRP concentration was within the reference range (<5.0 mg/L). The consistency of the subgroups for this particular study is shown in Table 4. The results shown in Table 5 demonstrate that fairly homogeneous values are obtained for any type of feeding in all subgroups of age, except for breastfed babies aged 3–5 days.

4

R. González-Tarancón et al. / Journal of Trace Elements in Medicine and Biology 39 (2017) 1–5

Table 4 Homogeneity of birth weight (g) between subgroups, expressed as median and 25 th and 75th percentiles, for the study of the effect of type of feeding. Only newborns with a birth weight >2500 g and hs-CRP concentrations <5.0 mg/L were selected. Differences between subgroups are assessed by U-Mann-Whitney or Kruskall-Wallis tests (significant p values <0.05). Breast

Mixed

Artificial

Birth weight (g)

n

Median

P25

P75

n

Median

P25

P75

n

Median

P25

P75

pb

Days of life 0–2 3–5 6–12 pa

28 44 24 0.065

3163 3110 3350

3005 2765 2955

3655 3400 3533

10 10 0 0.122

3050 3150 –

2740 3010 –

3220 3750 –

24 10 6 0.056

3190 3090 3250

2850 2790 2950

3537 3530 3400

0.065 0.082 0.594

a b

p value for differences in birth weight between days of life for each type of feeding group. p value for differences in birth weight between type of feeding for each days of life group.

Table 5 Effect of days of life at extraction and type of feeding on serum cooper concentrations (␮g/L), considering only newborns >2500 g and whose hs-CRP concentration was <5 mg/L. Differences between subgroups are assessed by U-Mann-Whitney or Kruskall-Wallis tests. Cu (␮g/L)

Breast

Days of life 0–2 3–5 6–12 pa a b *

Mixed

Artificial

n

Median

P2.5

P97.5

n

Median

P2.5

P97.5

n

Median

P2.5

P97.5

pb

28 44 24 0.004*

460 670 460

190 290 180

770 1060 1010

10 10 0 0.879

430 410 –

330 220 –

320 710 –

24 10 6 0.599

455 450 445

170 440 350

610 650 540

0.482 0.016* 0.243

p value for differences in Cu values between days of life for each type of feeding group. p value for differences in Cu values between type of feeding for each days of life group. significant p value <0.05.

1400 1200

Cu g/L

1000 800 600 400 200 0 1.0

0.5

0

0.5

1.0

1.5

2.0

Log hs-CRP (mg/L)

Fig. 1. Serum Cu concentrations (␮g/L) vs Log hs-CRP (mg/L) for newborns >2500 g and between 0 and 2 days of life.

4. Discussion Early detection of low serum Cu (and Cp) concentrations in newborns may be useful for identifying a deficient Cu intake or genetic diseases affecting Cu metabolism. Unfortunately, individuals belonging to this population show a wide range of serum Cu (and Cp) concentrations. Physiologically low concentrations are seen in many babies, while they can also be increased associated with some conditions [6–11]. For this reason, new Cu markers have been proposed over the last years that may help, for example, with an earlier diagnosis of WD. These are, for example, the measurement of the ultrafiltrable and/or the exchangeable Cu in serum [16], the enzymatic measurement of Cu non-bound to Cp [17], the establishment of the Cu isotopic ratio (65 Cu/63 Cu) in both serum and urine [18–20] or the quantification of Cp specific peptides from dried blood spots samples digested by trypsin [21]. However, these new potential markers have yet to be validated and introduced in clinical practice. Therefore, in spite of the limitations of serum Cu (and Cp) con-

centrations as surrogate for Cu status [22], its measurement is still the most available means to evaluate this parameter in a clinical setting. In this regard, the results obtained in this work for serum Cu and Cp concentrations in newborns agree, in general terms, with previously published conclusions [23–26]. We have found that the most important factor influencing serum Cu concentrations in newborns is the presence of an inflammatory state. Cp, as an acute-phase protein, increases its concentration in serum with inflammation and, as a result, a concomitant elevation of serum Cu concentration is produced [3,10]. In our study, for example, almost all preterm newborns in the subgroup of hs-CRP >20 mg/L have serum Cu concentrations similar to their term peers (see Table 2). Despite the clear tendency shown in Table 2, the correlation between the individual serum Cu and hs-CRP concentrations for each newborn is rather poor, with a global Spearman’s correlation coefficient of 0.30. When only the subgroup of newborns with >2500 g and 0–2 days is considered (Fig. 1) this correlation coefficient improves, but only slightly (Spearman’s correlation coefficient of 0.41). Therefore, although serum Cu and Cp concentrations are clearly influenced by hs-CRP concentrations, other factors must have an effect. The first obvious factor showing an effect on Cu and Cp concentrations is prematurity. Cu is accumulated in the foetal liver mainly in the third trimester of gestation to prevent its deficiency after birth [6,8]. For this reason, and due to the immaturity of their liver, preterm low weight newborns often present very low serum Cu and Cp concentrations [6,7]. This is clearly shown in Table 2 (and Table 3 for Cp) when comparing concentrations between preterm and term infants within each hs-CRP group. This effect has already been described in previous works [4,27], where a continuous and consistent rise of Cu concentrations from weeks 26 to 42 was observed. Contrarily, days of life at the time of extraction and type of feeding have a more modest influence on serum Cu concentrations, with a slight increase for sample collection at 3–5 days of age in breastfed newborns. It has already been demonstrated that breastfed infants absorb more Cu, probably due to the lower casein content of human milk [8]. In our study, when we eliminated the effect of the con-

R. González-Tarancón et al. / Journal of Trace Elements in Medicine and Biology 39 (2017) 1–5

founders “prematurity” and “inflammatory state”, we found that breastfed term newborns (>2500 g) have greater serum Cu concentrations than those mixed or artificially fed (Table 5), but only for the group of 3–5 days of life. It is fair to admit at this point that this particular study has some limitations, as the amount of Cu intake by each baby could not be estimated and some subgroups have a very small number of cases (Tables 4 and 5). However our results are consistent with the observation that human milk has a greater Cu content in the 3–7 days after delivery, which progressively decrease thereafter [12]. Besides the variables studied in this work, other factors may also have an effect on serum Cu concentrations, such as the speed of thriving in preterm newborns [9] or Zn supplementation, both in mothers during pregnancy [12] or in artificial formulas provided to newborns [28,29]. Even specific conditions have been shown to increase serum Cu concentrations in newborns as, for example, in haemolytic jaundice [25]. For all the above, Cu status in newborns must be evaluated together with their demographic parameters and take into account concomitant conditions that may affect these concentrations. 5. Conclusion Serum Cu (and Cp) concentrations in newborns are affected by several variables, which demand a proper stratification. Prematurity and inflammation are the main altering factors and its knowledge is mandatory to properly interpret serum Cu and Cp concentrations in this population. Therefore, hs-CRP should always be assayed in parallel to Cu status. Conflict of interests All authors declare that they have no conflicting interests. References [1] C.L. Keen, J.Y. Uriu-Hare, S.N. Hawk, M.A. Jankowski, G.P. Daston, C.L. Kwik-Uribe, et al., Effect of copper deficiency on prenatal development and pregnancy outcome, Am. J. Clin. Nutr. 67 (1998) 1003S–1011S. [2] H. Kodama, C. Fujisawa, Copper metabolism and inherited copper transport disorders: molecular mechanisms, screening, and treatment, Metallomics 1 (2009) 42–52. ˜ M. Olivares, R. Uauy, M. Araya, Risks and benefits of [3] D. López de Romana, copper in light of new insights of copper homeostasis, J. Trace Elem. Med. Biol. 25 (2011) 3–13. [4] A.M. Sutton, A. Harvie, F. Cockburn, J. Farquharson, R.W. Logan, Copper deficiency in the preterm infant of very low birthweight. Four cases and a reference range for plasma copper, Arch. Dis. Child. 60 (1985) 644–651. [5] M.L. Marquardt, S.L. Done, M. Sandrock, W.E. Berdon, K.W. Feldman, Copper deficiency presenting as metabolic bone disease in extremely low birth weight, short-gut infants, Pediatrics 130 (2012) e695–e698. [6] D. Beshgetoor, M. Hambidge, Clinical conditions altering copper metabolism in humans, Am. J. Clin. Nutr. 67 (1998) 1017S–1021S. [7] A. Cordano, Clinical manifestations of nutritional copper deficiency in infants and children, Am. J. Clin. Nutr. 67 (1998) 1012S–1016S. [8] M. Olivares, B. Lönnerdal, S.A. Abrams, F. Pizarro, R. Uauy, Age and copper intake do not affect copper absorption, measured with the use of Cu-65 as a tracer, in young infants, Am. J. Clin. Nutr. 76 (2002) 641–645.

5

[9] J. Burns, J.S. Forsyth, C.R. Paterson, Factors associated with variation in plasma copper levels in preterm infants of very low birth weight, Eur. J. Pediatr. 152 (1993) 240–243. [10] V.Q. Bui, A.D. Stein, A.M. DiGirolamo, U. Ramakrishnan, R.C. Flores-Ayala, M. Ramirez-Zea, et al., Associations between serum C-reactive protein and serum zinc ferritin, and copper in Guatemalan School children, Biol. Trace Elem. Res. 148 (2012) 154–160. [11] S.H. Hahn, Population screening for Wilson’s disease, Ann. N.Y. Acad. Sci. 1315 (2014) 64–69. [12] C.S.M. de Figueiredo, D.B. Palhares, P. Melnikov, A.J. da Cruz Montes Moura, S.C. dos Santos, Zinc and copper concentrations in human preterm milk, Biol. Trace Elem. Res. 136 (2009) 1–7. [13] J.G. Dorea, Iron and copper in human milk, Nutrition 16 (2000) 209–220. [14] A.A. Almeida, J.F.L.C. Lima, Optimized conditions and analytical performance for the determination of Cu in serum and urine samples using a single GFAAS procedure, At. Spectrosc. 22 (2001) 324–330. [15] P.R.M. Correia, E. Oliveira, P.V. Oliveira, Minimalism approach for determination of Cu: Fe and Zn in serum by simultaneous electrothermal atomic absorption spectrometry, Anal. Chim. Acta 458 (2002) 321–329. [16] S. El Balkhi, J. Poupon, J.-M. Trocello, A. Leyendecker, F. Massicot, M. Galliot-Guilley, et al., Determination of ultrafiltrable and exchangeable copper in plasma: stability and reference values in healthy subjects, Anal. Bioanal. Chem. 394 (2009) 1477–1484. [17] R. Squitti, Copper subtype of Alzheimer’s disease (AD): meta-analyses, genetic studies and predictive value of non-ceruloplasmim copper in mild cognitive impairment conversion to full AD, J. Trace Elem. Med. Biol. 28 (2014) 482–485. [18] M. Resano, M. Aramendía, L. Rello, M.L. Calvo, S. Bérail, C. Pécheyran, Direct determination of Cu isotope ratios in dried urine spots by means of fs-LA-MC-ICPMS. Potential to diagnose Wilson’s disease, J. Anal. At. Spectrom. 28 (2013) 98–106. [19] M. Aramendía, L. Rello, M. Resano, F. Vanhaecke, Isotopic analysis of Cu in serum samples for diagnosis of Wilson’s disease: a pilot study, J. Anal. At. Spectrom. 28 (2013) 675–681. [20] M. Aramendía, L. Rello, S. Bérail, A. Donnard, C. Pécheyran, M. Resano, Direct analysis of dried blood spots by femtosecond-laser ablation-inductively coupled plasma-mass spectrometry. Feasibility of split-flow laser ablation for simultaneous trace element and isotopic analysis, J. Anal. At. Spectrom. 30 (2015) 296–309. [21] A. deWilde, K. Sadilkova, M. Sadilek, V. Vasta, S.H. Hahn, Tryptic peptide analysis of ceruloplasmin in dried blood spots using liquid chromatography–tandem mass spectrometry: application to newborn screening, Clin. Chem. 54 (2008) 1961–1968. [22] R. Danzeisen, M. Araya, B. Harrison, C. Keen, M. Solioz, D. Thiele, et al., How reliable and robust are current biomarkers for copper status? Br. J. Nutr. 98 (2007) 676–683. [23] G. Lockitch, A.C. Halstead, L. Wadsworth, G. Quigley, L. Reston, B. Jacobson, Age- and sex-specific pediatric reference intervals and correlations for zinc, copper, selenium, iron, vitamins A and E, and related proteins, Clin. Chem. 34 (1988) 1625–1628. [24] A. Alimonti, F. Petrucci, S. Fioravanti, F. Laurenti, S. Caroli, Assessment of the content of selected trace elements in serum of term and pre-term newborns by inductively coupled plasma mass spectrometry, Anal. Chim. Acta 342 (1997) 75–81. [25] K.H. Schulpis, T. Karakonstantakis, S. Gavrili, C. Costalos, E. Roma, I. Papassotiriou, Serum copper is decreased in premature newborns and increased in newborns with hemolytic jaundice, Clin. Chem. 50 (2004) 1253–1256. [26] R. Abass, H. Hamdan, E. Elhassan, S. Hamdan, N. Ali, I. Adam, Zinc and copper levels in low birth weight deliveries in Medani Hospital, Sudan, BMC Res. Notes 7 (2014) 386. [27] A. Galinier, B. Périquet, W. Lambert, J. Garcia, C. Assouline, M. Rolland, et al., Reference range for micronutrients and nutritional marker proteins in cord blood of neonates appropriated for gestational ages, Early Hum. Dev. 81 (2005) 583–593. [28] C.J. Klein, Nutrient requirements for preterm infant formulas, J. Nutr. 132 (2002) 1395S–1577S. [29] I.J. Griffin, M. Domellöf, J. Bhatia, D.M. Anderson, N. Kler, Zinc and copper requirements in preterm infants: an examination of the current literature, Early Hum. Dev. 89 (2013) S29–S34.