| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH COMMUNICATIONS |
1 From the Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, Australia (LD); the Child Health Research Institute, based at Flinders Medical Centre and Women's and Children's Hospital (RAG and MM), School of Paediatrics and Reproductive Health (MM), and School of Agriculture, Food and Wine, (RAG), University of Adelaide, Adelaide, Australia; the University of Western Australia, Perth, Australia (KS); and the Nestlé Research Centre, Vers-chez-les Blanc, Lausanne, Switzerland (PVD)
2 Supported by Nestle, Switzerland. RAG and MM were supported by Senior Research Fellowships from the National Health Medical Research Council, Australia. The formula was provided by Nestle, Switzerland.
3 Reprints not available. Address correspondence to M Makrides, Child Nutrition Research Centre, Level 1, CRB, Women's and Children's Hospital, 72 King William Road, North Adelaide, SA 5006, Australia. E-mail: maria.makrides{at}cywhs.sa.gov.au.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objective: The objective was to evaluate the effect of fortifying infant formula (6 µg Se/L) with 2 concentrations of selenate (7 and 15 µg/L) on biochemical indicators of selenium status and growth at 16 wk in term infants.
Design: A randomized dose-response trial was conducted in 3 groups of term infants fed formula with different selenium concentrations [6 µg/L, F+0 (control); 13 µg/L, F+7; and 21 µg/L, F+15] and in a parallel breastfed reference group (BF; 11 ± 2 µg Se/L).
Results: One hundred sixty-one (47% males) infants completed the 16-wk study. Baseline plasma selenium was 0.3 ± 0.1 µmol/L. At 16 wk, plasma selenium had increased in all groups (P < 0.001) and was greater (P < 0.01) in the F+7 and F+15 groups and lower (P < 0.05) in the F+0 group than in the BF group. Plasma glutathione peroxidase increased in the F+15 group, decreased in the F+0 group, and, at 16 wk, was lower in the F+0 group than in the other groups (all P < 0.05). Erythrocyte selenium and glutathione peroxidase decreased in all groups (P < 0.05), but the magnitude of the change was greater in the F+0 than in the F+15 group (P < 0.05). There was no effect of selenium supplementation on growth.
Conclusions: Selenate fortification of formula resulted in an increase in plasma indicators of selenium status relative to indicators observed in infants fed low-selenium-containing formula. Although the erythrocyte indicators decreased in all groups, the 21-µg/L dose (F+15 group) resulted in a smaller decrease and in higher erythrocyte selenium than did the standard formula. Supplementation of low-selenium formula to provide a net selenium concentration close to that found in the breast milk of US women (18 µg/L) may be justified.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
35 selenoproteins (1-3). Selenoproteins require adequate selenium intake for synthesis and expression. Some selenoproteins, particularly glutathione peroxidase (GPx), are classically used in conjunction with plasma and erythrocyte selenium concentrations as indicators of selenium status (1, 2, 4, 5). Suboptimal selenium status is associated with a range of negative health outcomes including thyroid and immune dysfunction, viral infection, cardiovascular disease, inflammatory conditions, infertility, and an increased risk of some cancers (2, 5). The wide global variation in soil selenium is reflected in the selenium content of food and, hence, in adult and maternal selenium status, which correspond to the selenium concentration of newborns and human milk (1-3, 6). Infant formulas generally have endogenous selenium concentrations that are 30–50% lower than those in human milk from the same region (7-9), and formula feeding is associated with lower and declining indicators of selenium status compared with breastfeeding over the first few months of life (10-15). Overt selenium deficiency is manifested as Keshan disease, an endemic fatal cardiomyopathy. This condition is virtually unknown outside rural areas of China, where selenium concentrations in the soil and diet are extremely low. There have been a few reports of selenium deficiency associated with long-term parenteral nutrition (1). To date, no clinical outcomes have been clearly attributable to selenium deficiency or depletion in term infants (16, 17).
Organic selenomethionine is the predominant endogenous form of selenium in food, human milk, and infant formulas (1, 2, 5, 18). Inorganic selenite and selenate only occur in the diet through supplementation, have lower retention rates, and elicit different supplementation responses compared with organic forms (1, 2, 5). Most supplementation trials in term infants have used selenite (12, 13, 19-22), but one trial (14) added selenate to soy formula. These trials show considerable variation in the concentration and duration of supplementation and baseline infant selenium status. Compared with selenite, selenate is more stable, less pro-oxidative, and, in a short-term balance study, had similar net retention in term infants (23). However, selenate fortification of standard infant formulas has not been evaluated. The optimal form and dose of selenium supplementation required to achieve indicators of selenium status equivalent to those in breastfed infants remain unclear, particularly in newborns with relatively low concentrations.
This dose-response trial aimed to evaluate the effect of supplementing term formula with 2 concentrations of selenate on biochemical indicators of selenium status at 16 wk compared with the effect in a reference group of breastfed infants. This study was conducted in a region previously shown to have low plasma selenium concentrations in newborns (6).
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
37 wk gestational age, birth weight >2500 g and <4200 g) whose mothers had elected to exclusively formula feed from birth were eligible for the study. Any infants with major deformities or illness, including cardiovascular, gastrointestinal, renal, neurological, and metabolic disease, or participating in another clinical trial were excluded. The breastfed reference group had similar inclusion criteria, but were exclusively breastfed from birth. Participants were recruited from the Flinders Medical Centre, Adelaide, South Australia, in 2000 and 2001. The study was approved by the Flinders Clinical Research Ethics Committee.
Sample size calculations estimated that 35 infants per group were required to detect a 1.4-SD unpaired difference (power: 80%; 
= 0.05) (24) based on plasma selenium at 6 wk in breastfed (49 ± 12 µg/L) and formula-fed term (32 ± 11 µg/L) infants from our previous study (10) and allowing for 40% attrition. We aimed to enroll 60 breastfed infants to allow for a 50% rate of cessation of breastfeeding with 30 exclusively breastfed infants to complete the study.
Randomization and allocation to treatment group
Formula-fed infants were randomly assigned to 1 of 3 formula groups: control, unsupplemented formula (F+0), formula supplemented with 7 µg Se/L (F+7) as sodium selenate, or formula supplemented with 15 µg Se/L (F+15) as sodium selenate. Allocation occurred after consent by using opaque and sealed envelopes prepared according to a computer-generated randomization schedule, stratified by sex. Multiple births formed a single randomization unit, and allocation was according to the first-born twin. The formulas were manufactured in Switzerland (Nestlé Switzerland, Konolfingen) and provided in identical cans, distinguishable only by the color of the label. The codes were broken only after study completion and data analysis.
All test formulas were in powder form and were whey adapted (70:30, whey:casein) with a protein content of 1.8 g/100 kcal and standard nutritional composition providing 280 kJ/100 mL. The endogenous selenium concentration of the formulas was 6 µg/L. Sodium selenate was added by the manufacturer to provide total selenium concentrations approximately equivalent to Australian breast milk (13 µg/L) (6) or high-selenium breast milk (21 µg/L) (7). Target formula concentrations were confirmed by analysis. Feeding with the study formulas commenced following after baseline blood and urine collections at 3–5 d and continued ad libitum until the infant was 16 wk of age. Formula was provided in 465-g tins and prepared in the home according to the instructions on the label. The breastfed reference group was encouraged to exclusively breastfeed to 4 mo, and all participants were strongly encouraged to delay the introduction of solid food until 4 mo. Any nutrition provided in addition to breast milk or the study formula was recorded. Compliance (% of feeding) was determined from a 3-d record kept by the parents at home just before the final outcome assessment visit. Infants were defined as compliant if their total feeding was >80% of the allocated formula or breast milk.
Outcome assessments and data collection
At trial entry, maternal age, parity, smoking status, education, and supplement use were collected by interview. The indicators of selenium status assessed were selenium concentration and GPx activity in plasma and erythrocytes and urine selenium concentration. Baseline infant blood (by heel stick to coincide with neonatal screening) and urine samples were collected. Home visits were also conducted at 1–2, 4, 8, and 12 wk to measure weight (±10 g), length (±1 mm), and head circumference (± 1 mm) and to review feeding, supply formula, monitor adherence, and encourage retention. The final outcome assessment at 16 wk required a clinic visit to Flinders Medical Centre, at which time weight, length, and head circumference were measured, and 3 mL blood (by venipuncture) and urine samples (standard urine collection bags) were collected. Appropriate samples of expressed breast milk were collected by hand pump into sterile containers on each of 3 d immediately before the 16-wk visit; 10-mL aliquots were pooled and frozen.
Analytic procedures and sample processing
Plasma and erythrocytes were separated (15 min, 2000 x g, 4 °C) within 4 h, and plasma for selenium analysis was stored at –20 °C. Erythrocytes were washed 3 times with isotonic saline and reconstituted to the original blood volume to give an 
50% saline suspension, 25 µL of which was stored at –80 °C for hemoglobin determination. Equal volumes of lysis buffer (0.7 mmol/L β-mercaptoethanol and 2.7 mmol/L neutralized EDTA) were added to subsamples of the suspension, which were stored at –20 °C and –80 °C for erythrocyte selenium and GPx measurements, respectively. Similarly equal volumes of plasma and lysis buffer were stored at –80 °C for GPx determination. Frozen breast milk samples collected by mothers were thawed and mixed thoroughly, and 10-mL samples from each day were pooled and stored at –20 °C. Urine samples were stored in acid-washed containers at –20 °C.
Assay procedures
Selenium in plasma and erythrocytes was measured in quadruplicate with the Varian Spectra AA40 Carbon Furnace (Varian Australia Pty, Melbourne, Australia) using pyrolytically coated graphite partition tubes and Zeeman background correction at 196.0 nm. The method was calibrated using standard additions on the appropriate sample type, with 3 additions being used in the calibration curve. Palladium was used as a matrix modifier (25). Analysis of a quality-control serum sample (batch no. 98228A; QC Quality Control Technologies Pty Ltd, Charlestown, NSW) with a target value of 0.40 µmol/L (95% CI: 0.30, 0.50 µmol/L) gave a mean of 0.32 µmol/L and a between-day CV of 21% derived from duplicate rather than quadruplicate determinations used for the samples. On a similar basis, erythrocyte analysis yielded a mean of 0.40 µmol/L and a between-day CV of 14%. Plasma selenium is expressed in µmol/L and erythrocyte selenium as nmol/g hemoglobin of the saline suspension to adjust for any variance as a result of altered hematocrit in the heel-stick (baseline) and venipuncture (16 wk) blood samples.
Urine and breast milk samples were analyzed by inductively coupled plasma mass spectrometry following a 1:11 dilution with 0.2% HNO3. Analysis of standard quality-control materials yielded 93–101% of certified target values, and the limit of detection was 0.05 µmol/L. Creatinine was determined by the routine diagnostic method using the Beckman Sybchron CX3 (Beckman Coulter Australian Pty Ltd, Gladsville, NSW, Australia) and colorimetric alkaline picrate method. Urinary selenium was expressed in nmol/µmol creatinine and breast milk selenium in µmol/L.
The selenium content of the formula was determined by continuous-flow hydride generation atomic spectrometry with the use of a Varian AA-400 (Mulgrave, Australia) equipped with a PSC-56 autosampler and a VGA-57 hydride generation system as reported previously (26); the samples were acid digested with nitric acid and perchloric acid, which was followed by a hydrochloric acid reduction before selenium analysis. The accuracy of the selenium analysis was verified against NIST 1549 Non-Fat Milk Powder (NIST, Gaithersburg, MD) standard reference material. For ease of presentation, the selenium concentration of formula and breast milk is expressed in µg/L and can be converted to µmol by multiplying by 79.
Erythrocyte and plasma GPx activity in 1:200 and 1:10 dilutions of the lysis buffer, respectively, were determined by autoanalysis on a Cobas Bio Centrifugal Analyzer (Roche Diagnostics Australia Pty Ltd, Castle Hill, NSW, Australia) at 37 °C with t-butyl hydroperoxide as the substrate, as was described in detail elsewhere (4, 10). The within- and between-run CVs for infant quality-control samples were 3.7% and 6.6% for plasma and 5.6% and 8.8% for erythrocyte GPx activity. Enzyme activity was expressed in IU/g hemoglobin, where IUs represent micromoles of NADH (NADPH) oxidized per minute.
Statistical analysis
All analyses were performed by using SPSS for WINDOWS 13.0 (SPSS Inc, Chicago, IL). Summary data are expressed as means ± SDs unless stated otherwise. Different alphabetic or numeric scripts denote a significant difference (P < 0.05).
Differences in anthropometric measures and in indicators of selenium status between groups at time points were determined with ANOVA with Tukey's honestly significant difference multiple comparisons post hoc analysis. A linear mixed model including time, group, and their interaction with Bonferroni correction for post hoc multiple comparison was used to determine whether there was a significant time effect (change) in variables between baseline and 16 wk. Variables were transformed (square root) to satisfy assumptions of the mixed-effect model (normal distribution and equality of variance of residuals). The magnitude of changes between baseline and 4 mo (16 wk –days 3–5; positive value indicates an increase) were analyzed with a Kruskal-Wallis test with post hoc differences determined with a Mann-Whitney U test with the Holm method for adjusting for multiple comparisons (27). Correlations were determined by using nonparametric Spearman's rho. Significant differences were identified at an level of 0.05 (adjusted where appropriate). All available data were used in an intention-to-treat analysis.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
More than 95% of the infants were white. More formula-fed than breastfed infants were born by elective cesarean delivery (28% compared with 4%). Eight (14%) and 46 (41%) of the breastfeeding and formula feeding mothers, respectively, smoked during their pregnancy. For the breastfed and formula-fed infants, respectively, maternal age was 30 ± 6 and 29 ± 5 y and the duration of secondary schooling was 11.4 ± 0.7 and 10.8 ± 1.0 y. These characteristics did not differ between formula groups. There were 2 sets of formula-fed twins. Eighty-nine percent of formula-fed and 74% of breastfed infants were considered compliant.
Anthropometric data and age at baseline and outcome assessment are shown in Table 1
. Of the randomized infants, there was a small (<8%) but statistically significant difference in weight at study entry that was not evident at 16 wk (P = 0.086). There were no differences in length or head circumference at baseline or at 16 wk.
|
Baseline indicators of selenium status
There were no significant differences in baseline indicators between groups (Figure 1). The mean baseline indicators for all infants were as follows: plasma selenium, 0.33 ± 0.13 µmol/L (n = 149); erythrocyte selenium, 7.39 ± 3.41 nmol/g hemoglobin (n = 152); plasma GPx, 1128 ± 466 IU/L (n = 159); and erythrocyte GPx, 0.91 ± 0.62 IU/g hemoglobin (n = 154). The changes in indicators of selenium status over 16 wk are shown in Figure 1 (direction) and Table 2 (magnitude).
|
|
). At 16 wk, plasma selenium concentrations were equivalent in both selenium-supplemented groups and were higher than in the F+0 group (P < 0.001). Plasma selenium in the breastfed reference group was lower than that in the F+15 group (P = 0.001) but higher than that in the F+0 group (P = 0.013). There was no difference in the magnitude of the change between the supplemented groups, both of which had a significantly greater increase in plasma selenium than did the standard formula group (F+0) and the breastfed group. Breastfed infants also had a greater increase in plasma selenium than did the unsupplemented formula–fed group (F+0) (Table 2).
There was no difference in 16-wk plasma GPx concentrations between the supplemented groups and the breastfed reference group, but all of these groups had significantly higher plasma GPx activity than did the unsupplemented formula group (F+15, P = 0.012; F+7, P = 0.004; BF, P = 0.017) (Figure 1). Plasma GPx showed a strong group-by-time interaction (P = 0.006), with a significant increase in the F+15 group (P = 0.048) and a decrease in the F+0 group (P = 0.016). As expected, given the apparent different directions of the changes, the positive change in the F+15 group was not significantly different from the change in the F+7 group, but was significantly greater than the non-significant decrease in the BF group and the significant negative change in the F+0 group. Overall, the change in the F+7 group was greater than the significant decrease in the F+0 group but was not different from the change in the BF group.
Erythrocyte selenium and GPx
At 16 wk, erythrocyte selenium was higher in the F+15 group than in the F+0 group (P = 0.04). No other differences were evident. Erythrocyte selenium and GPx decreased significantly in all groups (Figure 1). For both erythrocyte indicators, the decrease in the F+15 group was less than that in the F+0 group but was not significantly different from that in the F+7 or the breastfed reference group (Figure 1, Table 2).
Relation between selenium concentration and GPx activity
When all infants were considered at baseline, there was a modest relation between selenium and GPx in both plasma (r = 0.20, P = 0.016; n = 144) and erythrocytes (r = 0.30, P < 0.001; n = 152). At 16 wk there were significant correlations between plasma selenium and GPx in formula-fed infants when considered as a whole (r = 0.29, P = 0.004; n = 96) but not in any of the feeding subgroups. However, in erythrocytes, selenium was associated with GPx in all formula-fed infants (r = 0.48, P < 0.001; n = 93), F+15 (r = 0.47, P = 0.005; n = 34), F+0 (r = 0.60, P < 0.001; n = 31) but not in the F+7 (r = 0.22, P = 0.26; n = 28) and breastfed (r = 0.33, P = 0.02; n = 50) groups.
Urinary selenium
At 16 wk there was a clear dose-response effect of selenium excretion in the formula-fed groups; urinary selenium in the F+15 group was 1.8 times that in the F+7 group, which in turn had a selenium concentration 2.3 times that of the F+0 group (Table 3). At week 16, log urinary selenium correlated with plasma selenium for all infants (r = 0.62, P < 0.001; n = 146) and in the breastfed (r = 0.64, P < 0.001; n = 49) and F+15 (r = 0.47, P < 0.01; n = 34) groups but not in the F+7 or F+0 group.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Differences in responses between plasma and erythrocyte indicators in infants were reported elsewhere (13, 20, 29), but the reasons are unclear. Plasma indicators reflect short term intake with plasma selenium most commonly used. Erythrocyte indicators reflect longer-term status associated with the availability of selenium at the time of cell synthesis (30). In plasma, selenium occurs predominantly as selenoprotein P (60–80%) and extracellular GPx3 (15%), whereas, in erythrocytes, selenium is found predominantly in cellular GPx1. These enzyme isoforms are different proteins (31, 32). There is evidence of preferential incorporation of dietary selenium into selenoproteins, with GPx1, GPx 3, and erythrocytes of lower priority (31, 33, 34). This may partly explain the differential response in plasma and erythrocyte indicators.
In adults, plasma GPx is a useful functional indicator of selenium status when plasma selenium is <70 µg/L. Above this concentration, GPx plateaus and is assumed to be saturated, which implies that intake is adequate and that requirements are being met (5, 30, 31). The validity of GPx as a functional indicator in infants has not been established. There are no recognized cutoffs for maximal GPx expression in infants, and requirements are based on the assumption that intakes and the status of breastfed infants are adequate (5). However, the selenium status of breastfed infants varies by geographic region. Because supplementation with 7 µg/L was sufficient to match the indicator profile of our Australian breastfed infants, and plasma GPx was not different between the F+7 and the breastfed groups, a possible interpretation is that plasma GPx in the F+7 group was saturated. This implies that intake is adequate and there is no case for supplementation at a higher dose. However, the response of plasma GPx (direction and magnitude) to the F+15 dose, the fall in erythrocyte GPx and the correlations between erythrocyte selenium and GPx in both formula-fed and breastfed infants suggest that GPx is not fully saturated. Supplementation to provide a net selenium concentration closer to that of US breast milk (18 µg/L), which is used as the basis for determination of the US infant selenium requirements (5), may be justified. The response of the erythrocyte indicators in the F+15 group, relative to infants fed standard formula, further supports the case for the higher dose.
In our infants with a low selenium status as newborns, supplementation was unable to reverse the decrease in erythrocyte indicators. New Zealand infants fed standard formula (7 µg/L) showed a response from 1 to 3 mo similar to that seen in our study for all indicators except plasma GPx, which did not change significantly (n = 11; P = 0.095) (19). In contrast with our data, supplemented formula (7 + 10 µg/L selenite) resulted in an increase in all 4 indicators (19). The reasons for this discrepancy are not clear, but they may be related to several factors, including differences in the assimilation of selenate and selenite. The 40% dropout rate in the New Zealand study may have resulted in a systematic error, and New Zealand infants may have an even lower newborn status than South Australian newborns (cord plasma: 36 µg Se/L compared with 50 µg Se/L (6)).
There appears to be a relation between newborn indicators of selenium status and the effect of dietary selenium. Infants with higher newborn plasma selenium maintained baseline plasma selenium and GPx and erythrocyte GPx values after feeding with breast milk, standard formula, and selenite-fortified formula, and net selenium concentrations were either much higher (21) or similar (13) to those in our study. Similar results were reported for infants fed soy formula with and without selenite (20, 21). In contrast, Smith et al (14) report that infants with high newborn plasma selenium (96 µg/L), fed over 16 wk with soy formula (2 µg/L) supplemented with selenate (11 µg/L) to a net concentration similar to that in our F+7 group our study, showed no increase in plasma selenium and a decrease in plasma GPx. These results also differ from the response of plasma indicators in our study, in which plasma selenium increased and plasma GPx was maintained over the same period, despite much lower newborn plasma selenium. Reasons for the disparity between the results from Smith et al's study (14) and data from our study and those mentioned above are not clear, but the use of both soy formula and selenate confounds direct comparisons. Erythrocyte selenium is only maintained (13, 21) or increased (21) in infants with high newborn plasma selenium fed human milk or selenite-supplemented formula with high selenium concentration (23 and 34 µg/L, respectively) (21) or as a result of maternal supplementation (13, 29).
Low selenium intakes may exacerbate low newborn stores (11). Temporal decreases in plasma GPx and in erythrocyte indicators are unlikely to be physiologic and are not seen in infants with high newborn plasma selenium fed high-selenium formula or breast milk (21, 35) and can be prevented or reversed in breastfed infants with maternal supplementation (13, 29). Hence, our data suggest that in regions where newborns have comparatively low selenium concentrations (6), supplementation of formula with selenium may be required. Our data also raise the question of whether some form of selenium supplementation of lactating women in these regions may also be indicated. This is supported by a study from Poland—a low selenium area. Infants exclusively fed low-selenium breast milk (9 µg/L) from 1 to 4 mo of age showed no change in blood or plasma selenium and a decrease in both plasma and erythrocyte GPx. Maternal supplementation with selenite or yeast increased the breast milk content to 14–16 µg/L and resulted in substantial increases in all indicators (29). No clinical or functional outcome data were provided.
Data regarding urinary selenium in infants are rare. In adults, the absorption of organic selenomethionine and selenate is high (>90%) (34), and one study has confirmed the latter in infants (23). The primary mechanism for selenium homeostasis is through urinary excretion rather than absorption (36, 37). Urinary selenium excretion is directly proportional to selenium intake and plasma selenium across a wide range of intakes (30). Our data show a clear dose-response relation between the selenate concentration in the formula and urinary selenium. In adults, urinary selenium losses increase with inorganic compared with organic intake (34, 38), but no infant data are available. Our results show that despite a similar net selenium concentration, the F+7 formula resulted in significantly higher urinary selenium concentrations than did breast milk, which suggests the form of selenium intake also effects urinary excretion in infants. Kumpulainen et al (13) report similar relative increased urinary losses in infants fed selenite-supplemented formula.
In conclusion, our biochemical data support the supplementation of formula with a low endogenous selenium content to a concentration at least equivalent to that of breast milk in the same geographical area, with no adverse effect on growth. Our data also raise the question of whether low-selenium-containing human milk provides an optimal profile of biochemical indicators of selenium status over the first few months of life in infants who have a correspondingly low newborn plasma selenium concentration and stores. This has theoretical implications for the use of breast milk as the appropriate basis for determining infant selenium requirements in the many comparatively low-selenium areas across the globe. It must be acknowledged that there is no evidence of functional or clinical consequences of feeding low-selenium breast milk and, hence, no reason to suggest an alternate approach to establishing infant selenium requirements. It should also be noted that, to our knowledge, there are no studies that have evaluated the impact of variation in breast milk selenium concentrations or in indicators of selenium status on short- or long-term health outcomes in healthy term infants. Given the increasingly varied and important role of selenium in health and well being (1, 2), there is a need to define the optimal intake in terms of biochemical and clinical outcomes for infants with variable newborn status.
|
ACKNOWLEDGMENTS |
|---|
The authors' responsibilities were as follows—LD, RAG, and MM (Steering Committee): monitored and managed the trial; MM: chaired the steering committee and supervised the clinical aspects of the trial; KS: provided medical support for the trial; LD and RAG: supervised the laboratory aspects of the trial; and LD: undertook the statistical analysis and wrote the manuscript with input from all coauthors. All authors contributed to the study design. PVD was an employee of Nestle. LD, RAG, KS, and MM had no financial interest in the production or sales of infant formula. Source document verification was done by a representative from Nestle. Data collection was supervised and managed, and the data were interpreted by the trial steering committee independent of Nestle.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, F+0; 6 µg Se/L) or formula fortified with 7 µg/L (
, F+7; 13 µg Se/L) or 15 µg/L (
, F+15; 21 µg Se/L) selenate. The change over 16 wk was assessed by using a linear mixed model (LMM) to analyze (Bonferroni adjustments for post hoc testing) the transformed variables (square root). Different superscripts indicate significant differences between groups at 16 wk, P < 0.05 (ANOVA with Tukey's post hoc testing). There were no significant differences in baseline indicators. Plasma selenium (LMM): group (df = 156, F = 9.6), time (df = 150, F = 269), and group-by-time interaction (df = 149, F = 11); significant increase in all groups (P < 0.001). Plasma selenium (ANOVA): F = 15.65, P < 0.001. Plasma glutathione peroxidase (LMM): group (df = 158, F = 3.2, P = 0.024), group-by-time interaction (df = 159, F = 4.3, P = 0.006); significant increase in the F+15 group (P = 0.048) and significant decrease in the F+0 group (P = 0.016). Plasma glutathione peroxidase (ANOVA): F = 4.87, P = 0.003. Erythrocyte selenium (LMM): time (df = 152, F = 138, P < 0.001), group-by-time interaction (df = 125, F = 3.9, P = 0.01); significant decrease in all groups: P < 0.001 (BF, F+7, and F+0), P = 0.002 (F+15). Erythrocyte selenium (ANOVA): F = 2.52, P = 0.06. Erythrocyte glutathione peroxidase (LMM): time (df = 155, F = 77.5, P = 0.001), group-by-time interaction (df = 155, F = 3.8, P = 0.012); significant decrease in all groups: P < 0.001 (BF and F+0), P = 0.035 (F+15), P = 0.001 (F+7). Erythrocyte glutathione peroxidase (ANOVA): F = 1.86, P = 0.139.