Home > Health & Research > Health Education Campaigns & Programs > Cochrane Neonatal Review > Selenium supplementation to prevent short-term morbidity in preterm neonates

Selenium supplementation to prevent short-term morbidity in preterm neonates

Skip sharing on social media links
Share this:

Authors

Brian A Darlow1, Nicola Austin2

Background - Methods - Results - Characteristics of Included Studies - References - Data Tables & Graphs


1Department of Paediatrics, Christchurch School of Medicine, CHRISTCHURCH, New Zealand [top]
2NICU, Christchurch Womens Hospital, Christchurch, New Zealand [top]

Citation example: Darlow BA, Austin N. Selenium supplementation to prevent short-term morbidity in preterm neonates. Cochrane Database of Systematic Reviews 2003, Issue 4. Art. No.: CD003312. DOI: 10.1002/14651858.CD003312.

Contact person

Brian A Darlow

Department of Paediatrics
Christchurch School of Medicine
PO Box 4345
CHRISTCHURCH
New Zealand

E-mail: brian.darlow@otago.ac.nz

Dates

Assessed as Up-to-date: 07 December 2010
Date of Search: 10 November 2010
Next Stage Expected: 07 December 2012
Protocol First Published: Issue 4, 2001
Review First Published: Issue 4, 2003
Last Citation Issue: Issue 4, 2003

What's new

Date / Event Description
07 December 2010
Updated

This review updates the existing review "Selenium supplementation to prevent short-term morbidity in preterm neonates" published in the Cochrane Database of Systematic Reviews (Darlow 2003).

Updated search found no new trials.

No changes to conclusions.

History

Date / Event Description
16 October 2008
Amended

Converted to new review format.

01 August 2003
New citation: conclusions changed

Substantive amendment

Abstract

Background

Selenium is an essential trace element and component of a number of selenoproteins including glutathione peroxidase, which has a role in protecting against oxidative damage. Selenium is also known to play a role in immunocompetence. Blood selenium concentrations in newborns are lower than those of their mothers and lower still in preterm infants. In experimental animals low selenium concentrations appear to increase susceptibility to oxidative lung disease. In very preterm infants low selenium concentrations have been associated with an increased risk of chronic neonatal lung disease and retinopathy of prematurity.

Objectives

To assess the benefits and harms of selenium supplementation in preterm or very low birth weight (VLBW) infants.

Search methods

Searches were made of the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 2, 2003), MEDLINE (1966 to May 2003), and Embase (1980 to May 2003). The reference lists of recent trials were also searched and abstracts from the Society for Pediatric Research from 1990 were hand-searched. This search was updated in November, 2010.

Selection criteria

Randomised controlled trials which compared selenium supplementation either parenterally or enterally with placebo or nothing from soon after birth in preterm or VLBW infants and which reported clinical outcomes were considered for the review.

Data collection and analysis

Data collection and analysis was conducted according to the standards of the Cochrane Neonatal Review Group.

Results

Three eligible trials were identified. Two trials, including one trial with a much larger sample size than the others combined, were from geographical areas with low population selenium concentrations. Meta-analysis of the pooled data showed a significant reduction in the proportion of infants having one or more episodes of sepsis associated with selenium supplementation [summary RR 0.73 (0.57 to 0.93); RD -0.10 (-0.17 to -0.02); NNT 10 (5.9 to 50)]. Supplementation with selenium was not associated with improved survival, a reduction in neonatal chronic lung disease or retinopathy of prematurity.

Authors' conclusions

Supplementing very preterm infants with selenium is associated with benefit in terms of a reduction in one or more episodes of sepsis. Supplementation was not associated with improved survival, a reduction in neonatal chronic lung disease or retinopathy of prematurity. Supplemental doses of selenium for infants on parenteral nutrition higher than those currently recommended may be beneficial. The data are dominated by one large trial from a country with low selenium concentrations and may not be readily translated to other populations.

Plain language summary

Selenium supplementation to prevent short-term morbidity in preterm neonates

Higher doses of selenium supplements may be able to reduce some complications for preterm babies, but more research is needed. Selenium is an essential trace element gained from nutrients. Babies are born with lower selenium concentrations in their blood than their mothers. In very preterm babies, low selenium is associated with an increased risk of complications. The review of trials of selenium supplementation for preterm babies found that it reduces sepsis (blood infection). It has not been shown to reduce other complications or increase survival. No adverse effects were reported. Higher than usual levels of selenium supplementation may be beneficial, but more research is needed as most of the evidence comes from a country where selenium levels were unusually low.

[top]

Background

Description of the condition

The trace element selenium is an essential component of a number of selenoproteins. These include the glutathione peroxidases (a family of enzymes that protect against oxidative injury by catalysing the breakdown of hydrogen peroxide and lipid hydroperoxides); iodothyronine deiodinase (which converts thyroxine [T4] to 3, 5, 3'-triodothyronine [T3]); and thioredoxin reductase (which is a key enzyme regulating the redox state of cells), (Arthur 1994; Rayman 2000). Selenium is also known to have a role in immunocompetence, selenium deficiency being associated with impairment of both cell-mediated immunity and B-cell function (Rayman 2000).

Description of the intervention

The consequences of low selenium concentrations are not fully known. A range of selenium deficiency diseases have been described in animals and in some selenium deficient geographical areas, such as New Zealand, selenium supplements are essential for maintaining animal health. During the late 1970s reports from endemically selenium deficient areas of China described two human diseases associated with severe nutritional selenium deficiency. However, both Keshan disease, a juvenile cardiomyopathy, and Kaschin-Beck disease, a chondrodystrophy, seem to require other causative co-factors in addition to selenium deficiency (Litov 1991; Rayman 2000). Outside of China, selenium deficiency diseases have been rarely recognised and then only following exceptional circumstances such as prolonged parenteral nutrition or severe malnutrition when selenium concentrations fall to < 0.20 micromol/L (16 micrograms/L). Clinical features have included muscle pain and tenderness, macrocytosis and pigmentary changes in hair and nails, and fatal cardiomyopathy (Litov 1991). The only reported adverse effect of low selenium levels in preterm infants has been increased erythrocyte fragility, which was accentuated by a diet high in polyunsaturated fatty acids and by iron supplements (Gross 1976).

Selenium levels in the soil vary considerably in different geographical locations with blood selenium concentrations in both animal and human populations reflecting these variations. Plasma selenium concentrations in newborn infants in all regions of the world are lower than those of their mothers (Litov 1991) and, in breast fed infants, then rise after birth. Preterm infants are born with slightly lower selenium and glutathione peroxidase concentrations than term infants, have low hepatic stores of selenium and, particularly if fed parenterally with solutions lacking selenium, these concentrations decline further in the first months of life (Lokitch 1989; Sluis 1992). In some preterm infants selenium concentrations may fall as low as 0.13 micromol/L (10 micrograms/L) and are amongst the lowest recorded in humans.

How the intervention might work

In experimental animals, selenium deficiency has been associated with increased susceptibility to oxidative lung injury (Hawker 1993; Kim 1991). Sick very preterm infants are exposed to many possible sources of oxygen radical products, including high concentrations of inspired oxygen, frequent alterations of blood flow to major organs, and inflammation with accumulation of neutrophils and macrophages. Low blood selenium concentrations in preterm infants have therefore been suggested as a potential risk factor for chronic neonatal lung disease (Amin 1980; Lokitch 1989) and retinopathy of prematurity (DeVoe 1988; Kretzer 1988). Darlow 1995 in a study from New Zealand, which has low soil and population selenium concentrations, reported that low selenium concentrations at 28 days were associated with an increased risk of adverse respiratory outcome in very low birthweight infants. However, uncertainty exists as to whether selenium supplementation in preterm infants will prevent such morbidity.

Objectives

Does selenium supplementation, given either enterally or parenterally, reduce the incidence of neonatal chronic lung disease, retinopathy of prematurity, or of late-onset sepsis in preterm or low birth weight infants, without causing clinically important side effects.

Secondary questions:

  1. Does the route of supplementation (enteral or parenteral) affect outcome?
  2. Does the dose of selenium supplementation affect outcome?
  3. Does the gestation or birthweight of the infant affect outcome?
  4. Does selenium supplementation lead to higher blood or plasma selenium concentrations at or beyond 28 days of age?

[top]

Methods

Criteria for considering studies for this review

Types of studies

Only randomised and quasi-randomised studies of the effect of selenium supplementation on one or more of: survival, chronic lung disease, bronchopulmonary dysplasia, retinopathy of prematurity and late-onset sepsis are included.

Types of participants

Preterm infants (defined as gestation less than 32 weeks) or infants with birthweight less than or equal to 2000 g.

Types of interventions

Selenium supplementation, in any formulation and either enterally or parenterally, versus no supplementation or placebo.

Types of outcome measures

Primary outcomes

Primary outcome measures: death before hospital discharge, neonatal chronic lung disease, defined as a) oxygen requirement at 28 days, b) oxygen requirement at 36 weeks post-menstrual age, retinopathy of prematurity (any and stages 3 or 4), one or more proven episodes of bacterial sepsis after the first week of life.

Secondary outcomes

Secondary outcomes measures: plasma or blood selenium and/or glutathione peroxidase concentrations at or beyond 28 days of age.

Side effects: skin eruptions, diarrhoea or other pre-defined clinical problems.

Search methods for identification of studies

The standard methods of the Neonatal Review Group of the Cochrane Collaboration were used. These involved searching the following databases: Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 2, 2003); MEDLINE from 1966 to May 2003 (using key words "selenium" and "chronic lung disease" or "bronchopulmonary dysplasia"; MeSH headings "infant, premature" or "infant, low birth weight"; and publication type "randomized controlled trial" or "controlled clinical trial"); and Embase from 1980 to May 2003. In addition, the reference lists of relevant articles were searched and abstracts from the Society for Pediatric Research from 1990 to May 2002 were also be hand-searched.

In November 2010, we updated the search as follows: MEDLINE (search via PubMed), CINAHL, EMBASE and CENTRAL (The Cochrane Library) were searched from 2003 to 2010. Search terms: (selenium AND (chronic lung disease or bronchopulmonary dysplasia)) AND ((infant, newborn[MeSH] OR newborn OR neon* OR neonate OR neonatal OR premature OR low birth weight OR VLBW OR LBW) AND (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized [tiab] OR placebo [tiab] OR clinical trials as topic [mesh: noexp] OR randomly [tiab] OR trial [ti]) NOT (animals [mh] NOT humans [mh]))

In November 2010 ClinicalTrials.gov and Controlled-Trials.com External Web Site Policy were also searched for relevant studies.

Data collection and analysis

Data collection and analysis was performed in accordance with the recommendations of the Cochrane Neonatal Review Group.

Selection of studies

Both reviewers examined all identified articles and the decision regarding inclusion/exclusion of the studies was by consensus. Where there were disagreements, the opinion of a third party was sought. Both reviewers completed data collection sheets, these were compared and any discrepancies were resolved by reference to the original sources.

Data extraction and management

Data was be collected on:

  • Supplementation dose;
  • Supplementation frequency;
  • Time of initiation of supplementation;
  • Route (enteral or parenteral) of supplementation;
  • Mean, standard deviation and range of gestational age in both supplemented and control groups;
  • Mean, standard deviation and range of birthweight in both supplemented and control groups;
  • Number of deaths in the supplemented and control groups;
  • Numbers developing neonatal chronic lung disease in the supplemented and control groups;
  • Numbers developing any and stage 3 or 4 retinopathy of prematurity in the supplemented and control groups;
  • Numbers with one or more proven episodes of bacterial sepsis after first week of life in the supplemented and control groups;
  • Blood or plasma selenium and glutathione peroxidase concentrations pre-randomisation in supplemented and control groups;
  • Blood or plasma selenium and glutathione peroxidase concentrations at or close to 28 days of age in supplemented and control groups;
  • Numbers developing skin eruptions, diarrhoea or other pre-defined side effects in supplemented and control group.

The review authors extracted data independently. Differences were resolved by discussion.

Assessment of risk of bias in included studies

The methodological quality of the included studies were assessed using the following key criteria: allocation concealment (blinding of randomisation), blinding of intervention, completeness of follow-up, and blinding of outcome measurement/assessment. For each criterion, assessment was yes, no, can't tell. Two review authors separately assessed each study. Any disagreement was resolved by discussion. This information was added to the Characteristics of Included Studies table.

In addition, for the update in 2010, the following issues were evaluated and entered into the Risk of Bias table:

  1. Sequence generation (checking for possible selection bias). Was the allocation sequence adequately generated?For each included study, we categorized the method used to generate the allocation sequence as:
    • adequate (any truly random process e.g. random number table; computer random number generator);
    • inadequate (any non random process e.g. odd or even date of birth; hospital or clinic record number);
    • unclear.
  2. Allocation concealment (checking for possible selection bias). Was allocation adequately concealed? For each included study, we categorized the method used to conceal the allocation sequence as:
    • adequate (e.g. telephone or central randomisation; consecutively numbered sealed opaque envelopes);
    • inadequate (open random allocation; unsealed or non-opaque envelopes, alternation; date of birth);
    • unclear.
  3. Blinding (checking for possible performance bias). Was knowledge of the allocated intervention adequately prevented during the study? At study entry? At the time of outcome assessment? For each included study, we categorized the methods used to blind study participants and personnel from knowledge of which intervention a participant received. Blinding was assessed separately for different outcomes or classes of outcomes. We categorized the methods as:
    • adequate, inadequate or unclear for participants;
    • adequate, inadequate or unclear for personnel;
    • adequate, inadequate or unclear for outcome assessors.
      In some situations there may be partial blinding e.g. where outcomes are self-reported by unblinded participants but they are recorded by blinded personnel without knowledge of group assignment. Where needed “partial” was added to the list of options for assessing quality of blinding.
  4. Incomplete outcome data (checking for possible attrition bias through withdrawals, dropouts, protocol deviations). Were incomplete outcome data adequately addressed? For each included study and for each outcome, we described the completeness of data including attrition and exclusions from the analysis. We noted whether attrition and exclusions were reported, the numbers included in the analysis at each stage (compared with the total randomised participants), reasons for attrition or exclusion where reported, and whether missing data were balanced across groups or were related to outcomes.Where sufficient information was reported or supplied by the trial authors, we re-included missing data in the analyses. We categorized the methods as:
    • adequate (< 20% missing data);
    • inadequate (greater than/or equal to 20% missing data):
    • unclear.
  5. Selective reporting bias. Are reports of the study free of suggestion of selective outcome reporting? For each included study, we described how we investigated the possibility of selective outcome reporting bias and what we found. We assessed the methods as:
    • adequate (where it is clear that all of the study’s pre-specified outcomes and all expected outcomes of interest to the review have been reported);
    • inadequate (where not all the study’s pre-specified outcomes have been reported; one or more reported primary outcomes were not pre-specified; outcomes of interest are reported incompletely and so cannot be used; study fails to include results of a key outcome that would have been expected to have been reported);
    • unclear.
  6. Other sources of bias. Was the study apparently free of other problems that could put it at a high risk of bias?
    For each included study, we described any important concerns we had about other possible sources of bias (for example, whether there was a potential source of bias related to the specific study design or whether the trial was stopped early due to some data-dependent process). We assessed whether each study was free of other problems that could put it at risk of bias as:
    • yes;no;or unclear.

If needed, we planned to explore the impact of the level of bias through undertaking sensitivity analyses.

Measures of treatment effect

Analysis of individual trials: For continuous variables such as duration of oxygen therapy, mean differences, and 95% confidence intervals were to be reported. For categorical outcomes such as mortality, the relative risks (RR) and 95% confidence intervals were to be reported.

Assessment of heterogeneity

We estimated the treatment effects of individual trials and examined heterogeneity between trials by inspecting the forest plots and quantifying the impact of heterogeneity using the I2 statistic. If we detected statistical heterogeneity, we planned to explore the possible causes (for example, differences in study quality, participants, intervention regimens, or outcome assessments) using post hoc subgroup analyses. We used a fixed effects model for meta-analysis.

Data synthesis

Separate analyses were conducted for each of the outcomes: death, neonatal chronic lung disease, retinopathy of prematurity (any and stage 3 or 4) and one or more proven episodes of sepsis. All analyses were conducted on an intention to treat basis. The data were analysed using the standard method of the neonatal review group using a fixed effect model, with use of relative risk, risk reduction, number needed to treat and their 95% confidence intervals.

Subgroup analysis and investigation of heterogeneity

If data were available, we planned subgroup analyses to address the following secondary questions:

  1. Does the route of supplementation (enteral or parenteral) affect outcome?
  2. Does the dose of selenium supplementation affect outcome?
  3. Does the gestation or birthweight of the infant affect outcome?
  4. Does selenium supplementation lead to higher blood or plasma selenium concentrations at or beyond 28 days of age?

[top]

Results

Description of studies

Nine potentially relevant trials were identified of which three met eligibility criteria and were included (Daniels 1996; Darlow 2000; Huston 1991). The report by Daniels 1996 gives in full data published in abstract form in Daniels 1995, and the report by Darlow 2000 gives in full data published in abstract form in Darlow 1998. Six trials (Bogye 1998a; Bogye 1998b; Ehrenkranz 1991; Rudolph 1981; Smith 1991; Tyrala 1996) were excluded from the final analysis because of inadequate methodology and/or because no clinical outcome data were reported. Data on selenium concentrations have not been reported in a manner allowing comparisons between studies and meta-analysis of these data have not been included in this recent version of the review.

Participants

The three included studies reported outcomes on 297 infants receiving selenium supplements and 290 control infants. One study (Darlow 2000) was much the largest and included 85% of all infants. This study was a multicentre study involving 8 New Zealand neonatal units and included infants with birthweight < 1500g admitted to a participating unit by 72 hours of age and without major abnormalities. Five hundred and thirty-four infants were randomised but five infants were withdrawn from the study, three by parents and two by the neonatal team caring for the infants and no data are available for these infants. Daniels 1996 included infants with birthweight < 2000g, who were expected to have parenteral nutrition for more than 5 days and no major congenital anomalies, liver or renal disease. Forty-four infants were enrolled but six excluded, including two who died (day 3 and day 11), three who received parenteral nutrition for less than 6 days and one because of protocol violations. Data, including group assignment, are not available for these infants. in addition, two further infants died before hospital discharge and again group assignment for these infants is not available. The study by Huston 1991 included infants with birthweight < 1000g with no congenital, metabolic or chronic white blood cell disease.

Interventions

The study by Daniels 1996 gave 3 micrograms/Kg/d selenious acid in parenteral nutrition, and received for a mean 18 days, in the treatment group versus nothing in controls. The study by Darlow 2000 gave 7 micrograms/Kg/d of sodium selenate added to parenteral nutrition in the treatment group versus nothing in controls, and 5 micrograms/Kg/d of sodium selenite in the treatment group or an equivalent volume of sterile water for control infants, when fed orally until 28 days of age. The study by Huston 1991 gave 1.5 micrograms/Kg/d of selenious acid added to parenteral nutrition, and received for a mean 38 days, in the treatment group versus nothing in controls.

Outcome measures

Daniels 1996 reported CLD (an oxygen requirement plus chest Xray changes) at 28 days, and one or more episodes of sepsis (defined as microbiologically confirmed or requiring antibiotics for at least five days). Darlow 2000 reported deaths pre-hospital discharge, CLD (an oxygen requirement) at 28 days and at 36 weeks post-menstrual age, any ROP in infants who were examined (ROP screening in New Zealand being routinely carried out in infants < 31 weeks gestation or < 1250g birthweight), and one or more episodes of sepsis (defined as clinical sepsis and a positive culture from either blood or cerebrospinal fluid) after the first week of life. Huston 1991 reported CLD at a mean 35 weeks post-menstrual age, any ROP, and one or more episodes of sepsis (not defined further).

Risk of bias in included studies

In the study by Daniels 1996, infants were stratified by birthweight < 1000 g and 1000 to 1999 g and randomisation was undertaken in pharmacy, although the method of randomisation is not stated. Investigators were blinded to group assignment. In the study by Darlow 2000, infants were stratified by hospital and by birthweight less than 1000 g and 1000 to 1499 g and randomisation was by telephone to the hospital pharmacy. This was the only study to give selenium supplementation orally after a period of parenteral nutrition (if required) and the investigators were blinded as to group assignment. This was also the only study to specify possible side effects of selenium supplementation; skin rashes, diarrhoea or a garlic odour on breath. In the study by Huston 1991, randomisation was by sealed envelopes and investigators were blinded to group assignment.

Effects of interventions

Primary outcomes:

Death pre-hospital discharge, Outcome 1.1

The studies by Darlow 2000 and Huston 1991 reported deaths pre-hospital discharge with no significant differences between selenium supplementation and control groups.

Oxygen use at 28 days in survivors, Outcome 1.2

The studies by Daniels 1996 and Darlow 2000 reported oxygen use at 28 days in survivors with neither showing a significant difference between groups. The pooled data also showed no significant difference [summary RR 0.99 (0.82 to 1.18); RD -0.01 (-0.09 to 0.08)].

Death or oxygen use at 28 days, Outcome 1.3

Only the study by Darlow 2000 reported death or oxygen use at 28 days with no significant difference between selenium supplementation and control groups.

Oxygen use at 36 weeks PMA in survivors, Outcome 1.4

The studies by Darlow 2000 and Huston 1991 reported oxygen use at 36 weeks post-menstrual age in surviving infants. Whilst there was no significant difference between groups in the study by Darlow 2000, the study by Huston 1991 did show a reduction in oxygen use associated with selenium supplementation of borderline statistical significance [ RR 0.33 (0.09 to 1.27); RD -0.40 (-0.79 to -0.01)]. The pooled data showed no significant difference between groups [summary RR 1.02 (0.75 to 1.39); RD 0.01 (-0.07 to 0.08)].

Retinopathy of prematurity (any grade) in examined infants, Outcome 1.5

The studies by Darlow 2000 and Huston 1991 reported on retinopathy of prematurity of any grade, with neither showing a significant difference between groups. The pooled data also showed no significant difference [summary RR 0.92 (0.71 to 1.18); RD -0.03 (-0.11 to 0.05)].

One or more episodes of sepsis, Outcome 1.6

All three studies reported on one or more episodes of sepsis and in Daniels 1996 and Darlow 2000 there was a significant reduction associated with selenium supplementation. The pooled data show a significant reduction in sepsis associated with selenium supplementation [summary RR 0.73 (0.57 to 0.93); RD -0.10 (-0.17 to -0.02); NNT 10 (5.9 to 50)], with no significant heterogeneity.

Side effects

Only the study by Darlow 2000 reported on side effects of selenium supplementation (skin rashes, diarrhoea or garlic odour on breath), and noted that none were recorded.

Secondary outcomes:

Blood or plasma selenium concentrations at or beyond 28 days of age.

Data on selenium concentrations have not been reported in a similar manner across the studies and do not allow a meta-analysis to be performed. The study by Huston 1991, which used the lowest dose of selenium supplementation added to parenteral nutrition (1.5 micrograms/Kg/d), found supplemented infants had significantly higher serum concentrations when oral feeds were commenced (mean day 14 to 15) than control infants, although these concentrations were below pre-randomisation concentrations. The study by Daniels 1996 used a dose of 3 micrograms/Kg/d added to parenteral nutrition and found plasma selenium concentrations fell compared with pre-randomisation concentrations by three weeks of age in control infants, whilst this decline was prevented by supplementation. By six weeks of age there was no significant difference between the groups with selenium concentrations in both being similar to pre-randomisation concentrations. The study by Darlow 2000 supplemented parenteral nutrition with 7 micrograms/Kg/d and enteral feeding with 5 micrograms/Kg/d of selenium. Supplemented infants had plasma selenium concentrations significantly higher (nearly two-fold) at 28 days and 36 weeks post-menstrual age than pre-randomisation concentrations and these were similar to concentrations found in healthy breast-fed term infants in this population. By contrast, selenium concentrations in control infants showed a non-significant fall at these times compared with pre-randomisation concentrations. Plasma glutathione peroxidase concentrations were also significantly higher at 28 days and 36 weeks post-menstrual age compared with control infants and with pre-randomisation concentrations.

The study by Darlow 2000 included 243 infants with birthweight < 1000 g and noted that confining analysis to this subgroup of infants did not reveal significant differences between the groups with respect to primary or secondary outcomes, however the data have not been reported.

Discussion

The study by Darlow 2000 was a large, multicentre randomised controlled trial with 80% power to detect a 12% reduction in oxygen dependency at 28 days. However, neither this study nor pooled data from this study and one of two smaller studies reporting similar outcomes showed any benefit from selenium supplementation with respect to deaths prior to discharge, oxygen use in survivors at 28 days or at 36 weeks post-menstrual age, or retinopathy of prematurity of any stage, with relative risks for all analyses being close to unity.

The study by Darlow 2000 reported that lower plasma selenium concentrations before randomisation were associated with an increased risk of adverse respiratory outcome at 28 days and a trend to increased risk of adverse outcome at 36 weeks post-menstrual age, after controlling for gestational age, antenatal steroids, CRIB score and hospital. There was a similar association between maternal selenium concentrations at enrolment and neonatal outcome. The relationship between pre-randomisation infant selenium concentrations and total days in oxygen was non-linear with a threshold of between 0.2 and 0.4 micromols/L below which, in this population, selenium concentrations had an effect on days of oxygen requirement. These findings suggest that if oxidative damage contributes to respiratory morbidity in preterm infants, and selenium has a role in preventing or ameliorating such damage, that it may be critical to achieve improved selenium status earlier than was the case in the studies included in this review. Low selenium concentrations at birth reflect low body stores (Bayliss 1985) and also correlate with maternal concentrations. At least in populations with low selenium status, the role of maternal selenium supplementation from perhaps 20 weeks' gestation could be the subject of further study, although given that very preterm infants constitute only around 1% of births, large numbers of pregnant women (at least 5, 000) would have to be recruited.

Two of the studies in this review come from geographical areas recognised as having low soil and population concentrations of selenium. There are good theoretical reasons to expect that, if selenium supplementation does result in decreased morbidity, this would be more readily apparent in studies from populations with low selenium concentrations. Nevertheless, the fact that the one large study is from such a population and that no large trials have been carried out in populations with higher concentrations, does mean that the results may not be generalisable.

The mean pre-randomisation plasma selenium concentrations in the study by Daniels 1996 (South Australia) were 0.34 and 0.36 micromols/L, in the study and control groups respectively, and in the study by Darlow 2000 (New Zealand) were 0.33 micromols/L in both groups. In contrast the mean pre-randomisation serum selenium concentrations in the study by Huston 1991 (Portland, Oregan) were 0.91 and 0.81 micromols/L in the study and control groups. Preterm infants generally have lower selenium concentrations than term infants in the same population, and in healthy term breast-fed infants concentrations then rise after birth. North American recommendations are for preterm infants to be supplemented with 2 micrograms/Kg/d selenium whilst receiving parenteral nutrition (Reifen 1993). The data from Huston 1991 show that a slightly lower dose (1.5 micrograms/Kg/d) resulted in serum concentrations still below starting concentrations after two weeks. In another study from North America, Ormsby 1998, reported that a dose of 4 micrograms/Kg/d added to parenteral nutrition for an average of more than three weeks did not maintain day one plasma selenium concentrations in preterm infants. Taken together, the studies of Daniels 1996 and Darlow 2000 suggest that 3 micrograms/Kg/d selenium supplementation in preterm infants may maintain cord concentrations but supplementation of 5-7 micrograms/Kg/d may be required to raise concentrations above those in cord blood to close to those found in healthy breast-fed term infants. Additional factors that may influence selenium concentrations in preterm infants include the form of selenium supplementation, route of administration and interactions with other nutrients, and the selenium content of enteral feeds.

Two studies, Daniels 1996 and Darlow 2000, as well as the pooled data showed a significant reduction in one or more episodes of sepsis associated with selenium supplementation, with the number needed to treat to prevent sepsis in one infant being 10. Selenium is known to have a role in immunocompetence (Rayman 2000). Neutrophils and macrophages from selenium deficient animals have low glutathione peroxidase concentrations, which may affect their antimicrobial properties, and animal studies have also shown that immunoglobulin antibody responses may be enhanced by selenium supplementation (Turner 1991). The relationship between selenium status and infections in preterm infants in different populations and the role of selenium supplementation is an important area for further research.

Limitations of this review

Two included studies did not report outcome for all participants. Daniels 1996 did not report any data, including group assignment, for six infants, two of whom died. In the study by Darlow 2000 examination for retinopathy of prematurity followed national guidelines (< 31 weeks gestation or < 1250g birthweight as a routine) and hence not all surviving infants were examined. Darlow 2000 also defined sepsis as episodes after the first week of life.

Two of the studies in this review, one of which accounts for 85% of participants, come from geographical areas recognised as having low soil and population concentrations of selenium. There are good theoretical reasons to expect that, if selenium supplementation does result in decreased morbidity, this would be more readily apparent in studies from populations with low selenium concentrations. Nevertheless, the fact that the one large study is from such a population and that no large trials have been carried out in populations with higher concentrations, does mean that the results may not be generalisable.

Authors' conclusions

Implications for practice

Supplementing very preterm infants with selenium is associated with benefit in terms of a reduction in one or more episodes of late-onset sepsis. Supplementation was not associated with improved survival, a reduction in neonatal chronic lung disease or retinopathy of prematurity. The data would suggest that the currently recommended dose of selenium for preterm infants receiving parenteral nutrition, 2 micrograms/Kg/d, is inadequate to maintain cord selenium concentrations, whilst doses of 3 micrograms/Kg/d may prevent a decline in cord levels and doses of up to 7 micrograms/Kg/d may be required to achieve concentrations above those in cords and close to concentrations found in healthy breast fed infants. These latter data come from New Zealand, a country with low soil and population selenium concentrations, and may not be readily translated to other populations. Beyond the reduction in late-onset sepsis, the benefits of selenium supplementation remain more theoretical than proven. However, the decline in selenium concentrations seen in non-supplemented very preterm infants is not physiological and, given that selenium is recognised as an essential trace element, should be avoided.

Implications for research

The impact of maternal selenium supplementation from 20 weeks gestation on the outcome for very preterm infants deserves investigation, although given that very preterm infants constitute only around 1% of births, large numbers of pregnant women would have to be recruited.

Further studies are warranted on the relationship between selenium status and infections in preterm infants.

Acknowledgements

We wish to thanks reviewers of the CNRG for helpful comments.

Editorial support of the Cochrane Neonatal Review Group has been funded with Federal funds from the Eunice Kennedy Shriver National Institute of Child Health and Human Development National Institutes of Health, Department of Health and Human Services, USA, under Contract No. HHSN267200603418C.

Contributions of authors

BD wrote the protocol. Both reviewers searched the databases, excerpted data, analysed results and wrote the review.

The November 2010 update was conducted centrally by the Cochrane Neonatal Review Group staff (Yolanda Montagne, Diane Haughton, and Roger Soll). This update was reviewed and approved by BD.

Declarations of interest

BD and NA are both authors of the largest trial included in this review.

Differences between protocol and review

  • None noted.

Additional tables

  • None noted.

Potential conflict of interest

  • None noted.

[top]

Characteristics of studies

Characteristics of Included Studies

Daniels 1996

Methods

Single centre randomised controlled trial
Blinding of randomisation: yes
Blinding of intervention: yes
Complete follow-up: no
Blinding of outcome measures: yes

Participants

N=38 [44 infants randomised but data only reported for 38 - see text]
Birthweight less than 2000g; stratified as less than 1000g or 1000-1999g
Expected to have TPN for more than 5 days
No major congenital anomalies, liver or renal disease

Interventions

19 infants randomised to treatment and 19 to control groups
3 microg/kg/d of selenious acid added to TPN for as long as this required (mean 18-19 days) vs no supplementation

Outcomes

CLD (defined as oxygen requirement at 28 days plus chest Xray changes
One or more episodes of sepsis (defined as microbiologically confirmed or antibiotics for at least 5 days)
Changes in mean plasma and erythrocyte selenium or glutathione peroxidase activity at 3 and 6 weeks

Notes

Total selenium intake from all sources monitored

Risk of bias table
Item Judgement Description
Adequate sequence generation? Unclear

Single centre randomised controlled trial

Allocation concealment? Yes

Blinding of randomisation: yes

Blinding? Yes

Blinding of intervention: yes
Blinding of outcome measures: yes

Incomplete outcome data addressed? No

Complete follow-up: no

Free of selective reporting? Unclear
Free of other bias? Unclear

Darlow 2000

Methods

Multicentre randomised controlled trial
Blinding of randomisation: yes
Blinding of intervention: yes
Complete follow-up: yes
Blinding of outcome measurement: yes

Participants

N=529 [534 infants randomised but data only reported for 529 - see text]
Birthweight less than 1500g, stratified by hospital and less than 1000g or 1000-1499g
No major anomalies

Interventions

268 infants randomised to treatment and 261 to control groups
7 microg/kg/d of sodium selenate added to TPN vs no supplementation when intravenously fed and 5 microg/kg/d sodium selenite vs equal volume (0.5 ml/Kg) sterile water when orally fed until 36 weeks PMA or discharge home

Outcomes

Death before hospital discharge
CLD defined as oxygen requirement at 28 days of age and at 36 weeks PMA
ROP of any stage in examined infants
One or more episodes of sepsis (defined as clinical sepsis and positive culture from blood or CSF) after 1 week of age
Plasma selenium and glutathione peroxidase concentrations at 28 days and 36 weeks PMA

Notes

Infants changed from parenteral to oral supplements when tolerating 3 ml/hr

Risk of bias table
Item Judgement Description
Adequate sequence generation? Unclear

Multicentre randomised controlled trial

Allocation concealment? Yes

Blinding of randomisation: yes

Blinding? Yes

Blinding of intervention: yes
Blinding of outcome measurement: yes

Incomplete outcome data addressed? Yes

Complete follow-up: yes

Free of selective reporting? Unclear
Free of other bias? Unclear

Huston 1991

Methods

Single centre randomised controlled trial
Blinding of randomisation: yes
Blinding of intervention: yes
Complete follow-up: yes
Blinding of outcome measurement: yes

Participants

N=20
Birthweight less than 1000g
No congenital metabolic or white blood cell diseases

Interventions

10 infants randomised to treatment and 10 to control groups
1.5 microg/kg/d of selenious acid added to TPN for as long as this required (mean 42 days) vs no supplementation

Outcomes

Death in hospital
CLD defined as oxygen requirement at 60 days (equivalent to 35.1 weeks PMA from mean gestation of 26.5 weeks)
ROP of any stage and grades III and IV in examined infants
One or more episodes of sepsis (not defined further)
Serum selenium and WBC glutathione peroxidase concentrations at time enteral feeding commenced (mean 14-15 days) and at time fully enterally fed (mean 42 days)

Notes

Also assessed serum levels of copper and zinc and WBC superoxide dismutase activity

Risk of bias table
Item Judgement Description
Adequate sequence generation? Unclear

Single centre randomised controlled trial

Allocation concealment? Yes

Blinding of randomisation: yes

Blinding? Yes

Blinding of intervention: yes
Blinding of outcome measurement: yes

Incomplete outcome data addressed? Yes

Complete follow-up: yes

Free of selective reporting? Unclear
Free of other bias? Unclear

Characteristics of excluded studies

Bogye 1998a

Reason for exclusion

No clinical outcome data
Serum selenium and glutathione concentrations at 14 days only
The Biofactors report is labelled an "Extended Abstract" and repeats data in primary report

Bogye 1998b

Reason for exclusion

No clinical outcome data
Serum selenium and glutathione concentrations at 14 days only
Report contains similar data to (includes 36 compared with 28 infants), and may overlap with, Bogye 1998a. The author is being contacted for clarification.

Ehrenkranz 1991

Reason for exclusion

No clinical outcome data
Main aim of study was to assess enteral selenium absorption and retention using a stable isotope

Rudolph 1981

Reason for exclusion

Not a trial of selenium supplementation

Smith 1991

Reason for exclusion

Oral supplementation begun at mean 15-26 days.
Data not reported for 25 of 71 enrolled infants.
No clinical outcome data.

Tyrala 1996

Reason for exclusion

Only infants with no evidence of disease process included.
Study infants not all feeding study formula exclusively until 4 weeks of age.
No clinical outcomes

[top]

References to studies

Included studies

Daniels 1996

* Daniels L, Gibson R, Simmer K. Randomised clinical trial of parenteral selenium supplementation in preterm infants. Archives of Disease in Childhood 1996;74:F158-64. [MEDLINE: 96273680; Other: EMBASE 1996153774]

Daniels LA, Gibson RA, Simmer K. Selenium (SE) supplementation of preterm infants. Journal of Paediatrics and Child Health 1995;31:A22.

Darlow 2000

Darlow BA, Inder TE, Sluis KB, Buss H, Graham P, Mogridge N, Winterbourn CC, The New Zealand Neonatal Study Group. Randomised controlled trial of selenium supplementation in New Zealand VLBW infants. Pediatric Research 1998;43:258A.

* Darlow BA, Winterbourn CC, Inder TE, Graham PJ, Harding JE, Weston PJ, Austin NC, Elder DE, Mogridge N, Buss IH, Sluis KB, The New Zealand Neonatal Study Group. The effect of selenium supplementation on outcome in very low birth weight infants: A randomized controlled trial. Journal of Pediatrics 2000;136:473-80. [MEDLINE: 10753245]

Huston 1991

Huston RK, Jelen BJ, Vidgoff J. Selenium supplementation in low-birthweight premature infants: Relationship to trace metals and antioxidant enzymes. J Parent Ent Nutr 1991;15:556-59. [MEDLINE: 92046632]

Excluded studies

Bogye 1998a

* Bogye G, Alfthan G, Machay T. Bioavailability of enteral yeast-selenium in preterm infants. Biological Trace Element Research 1998;65:143-51. [MEDLINE: 99095845]

Bogye G, Alfthan G, Machay T. Randomized clinical trial of enteral yeast-selenium supplementation in preterm infants. Biofactors 1998;8:139-42.

Bogye 1998b

Bogye G, Alfthan G, Machay T, Zubovics L. Enteral yeast-selenium supplementation in preterm infants. Archives of Disease in Childhood 1998;78:F225-26. [MEDLINE: 98378660]

Ehrenkranz 1991

Ehrenkranz RA, Gettner PA, Nelli CM, Sherwonit EA, Williams JE, Ting BT, Janghorbani M. Selenium absorption and retention by very-low-birth-weight infants: Studies with the extrinsic stable isotope tag 74Se. Journal of Pediatric Gastroenterology and Nutrition 1991;13:125-33. [MEDLINE: 92045183]

Rudolph 1981

Rudolph N, Preis O, Bitzos EI, Reale MM, Wong SL. Hematological and selenium status of low-birth-weight infants fed formulas with and without iron. Journal of Pediatrics 1981;99:57-62. [MEDLINE: 81241841; Other: EMBASE 1981206621]

Smith 1991

Smith AM, Chan GM, Moyer-Milieur LJ, Johnson CE, Gardner BR. Selenium status of preterm infants fed human milk, preterm formula, or selenium-supplemented preterm formula. Journal of Pediatrics 1991;119:429-33.

Tyrala 1996

Tyrala EE, Borschel MW, Jacobs JR. Selenate fortification of infant formulas improves the selenium status of preterm infants. American Journal of Clinical Nutrition 1996;64:860-65.

Studies awaiting classification

  • None noted.

Ongoing studies

  • None noted.

Other references

Additional references

Amin 1980

Amin S, Chen SY, Collipp PJ, Castro-Magan M, Maddaiah JVT, Klein SW. Selenium in premature infants. Nutrition and Metabolism 1980;24:331-40.

Arthur 1994

Arthur JR, Beckett GJ. New metabolic roles for selenium. Proceedings of the Nutrition Society 1994;53:615-24.

Bayliss 1985

Bayliss PA, Buchanan BE, Hancock RGV, Zlotkin SH. Tissue selenium accretion in premature and full-term infants and children. Biological Trace Element Research 1985;7:55-61.

Darlow 1995

Darlow BA, Inder TE, Graham PJ, Sluis KB, Malpas TJ, Taylor BJ, Winterbourn CC. The relationship of selenium status to respiratory outcome in the very low birth weight infant. Pediatrics 1995;96:314-9.

DeVoe 1988

DeVoe WM. Prevention of retinopathy of prematurity. Seminars in Perinatology 1988;12:373-80.

Gross 1976

Gross S. Hemolytic anaemia in premature infants: relationship to vitamin E, glutathione peroxidase and erythrocyte lipids. Seminars in Hematology 1976;13:187-99.

Hawker 1993

Hawker FH, Ward HE, Stewart PM, Wynne LA, Snitch PJ. Selenium deficiency augments the pulmonary toxic effects of oxygen exposure in the rat. European Respiratory Journal 1993;6:1317-23.

Kim 1991

Kim HY, Picciano MF, Wallig MA, Milner JA. The role of selenium nutrition in the development of neonatal rat lung. Pediatric Research 1991;29:440-45.

Kretzer 1988

Kretzer FL, Hittner HM. Retinopathy of prematurity: clinical implications of retinal development. Archives of Disease in Childhood 1988;63:1151-67.

Litov 1991

Litov RE, Combs GF. Selenium in pediatric nutrition. Pediatrics 1991;87:339-51.

Lokitch 1989

Lockitch G, Jacobson B, Quigley G, Dison P, Pendray M. Selenium deficiency in low birth weight neonates: an unrecognized problem. Journal of Pediatrics 1989;114:865-70.

Ormsby 1998

Ormsby AR, Tyrala EE. Se sufficiency is not achieved with currently recommended dosages of IV Se intake. Pediatric Research 1998;43:268A.

Rayman 2000

Rayman MP. The importance of selenium to human health. Lancet 2000;356:233-41.

Reifen 1993

Reifen RM, Zlotkin S. Microminerals. In: Tsang RC, Lucas A, Uauy R, Zlotkin S, editor(s). Nutritional needs of the preterm infant: scientific basis and practical guidelines. Baltimore: Williams and Wilkins, 1993:195-207.

Sluis 1992

Sluis KB, Darlow BA, George PM, Mogridge N, Dolamore BA, Winterbourn CC. Selenium and glutathione peroxidase levels in premature infants in a low selenium community (Christchurch, New Zealand). Pediatric Research 1992;32:189-94.

Turner 1991

Turner RJ, Finch JM. Selenium and the immune response. Proceedings of the Nutrition Society 1991;50:275-85.

Other published versions of this review

Darlow 2003

Darlow BA, Austin N. Selenium supplementation to prevent short-term morbidity in preterm neonates. Cochrane Database of Systematic Reviews 2003, Issue 4. Art. No.: CD003312. DOI: 10.1002/14651858.CD003312.

[top]

Data and analyses

1 Supplemental selenium vs placebo or nothing

Outcome or Subgroup Studies Participants Statistical Method Effect Estimate
1.1 Death pre-hospital discharge 2 549 Risk Ratio (M-H, Fixed, 95% CI) 0.92 [0.48, 1.75]
1.2 Oxygen use at 28 days in survivors 2 548 Risk Ratio (M-H, Fixed, 95% CI) 0.99 [0.82, 1.18]
1.3 Death or oxygen use at 28 days 1 529 Risk Ratio (M-H, Fixed, 95% CI) 0.97 [0.80, 1.16]
1.4 Oxygen use at 36 weeks PMA in survivors 2 521 Risk Ratio (M-H, Fixed, 95% CI) 1.02 [0.75, 1.39]
1.5 Retinopathy of prematurity (any grade) in examined infants 2 466 Risk Ratio (M-H, Fixed, 95% CI) 0.92 [0.71, 1.18]
1.6 One or more episodes of sepsis 3 583 Risk Ratio (M-H, Fixed, 95% CI) 0.73 [0.57, 0.93]

[top]

Figures

  • None noted.

Sources of support

Internal sources

  • No sources of support provided.

External sources

  • No sources of support provided.

This review is published as a Cochrane review in The Cochrane Library, Issue 2, 2011 (see http://www.thecochranelibrary.com External Web Site Policy for information). Cochrane reviews are regularly updated as new evidence emerges and in response to feedback. The Cochrane Library should be consulted for the most recent recent version of the review.