0
Original Research |

Effect of Supplementation With High-Selenium Yeast on Plasma Lipids: A Randomized Trial FREE

Margaret P. Rayman, DPhil; Saverio Stranges, MD, PhD; Bruce A. Griffin, PhD; Roberto Pastor-Barriuso, PhD; and Eliseo Guallar, MD, DrPH
[+] Article and Author Information

From the University of Surrey, Guildford, United Kingdom; Health Sciences Research Institute, University of Warwick Medical School, Coventry, United Kingdom; National Center for Epidemiology, Carlos III Institute of Health, Madrid, Spain; and Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland.


Acknowledgment: The authors acknowledge the support of Cancer Research UK (formerly Cancer Research Campaign) for the UK PRECISE Pilot Study. A contribution toward funding for the lipid measurements was provided by the Cardiovascular Disease, Diabetes & Metabolism Research Theme, University of Surrey. The authors also thank Max Wong, PhD, who carried out the lipid analyses, and the personnel of the Medical Research Council General Practice Research Framework and the Clinical Trials and Statistics Unit, Institute of Cancer Research.

Grant Support: By Cancer Research UK (formerly Cancer Research Campaign) and the Cardiovascular Disease, Diabetes & Metabolism Research Theme, University of Surrey.

Potential Conflicts of Interest: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M10-1758.

Reproducible Research Statement:Study protocol and data set: Available from Dr. Rayman (e-mail, m.rayman@surrey.ac.uk). Statistical code: Available from Dr. Guallar (e-mail, eguallar@jhsph.edu).

Requests for Single Reprints: Margaret P. Rayman, DPhil, Faculty of Health and Medical Sciences, University of Surrey, Guildford GU2 7XH, United Kingdom; e-mail, m.rayman@surrey.ac.uk.

Current Author Addresses: Drs. Rayman and Griffin: Faculty of Health and Medical Sciences, University of Surrey, Guildford GU2 7XH, United Kingdom.

Dr. Stranges: Health Sciences Research Institute, University of Warwick Medical School, Medical School Building, Room A105, Gibbet Hill Campus, Coventry CV4 7AL, United Kingdom.

Dr. Pastor-Barriuso: National Center for Epidemiology, Carlos III Institute of Health, Sinesio Delgado 6, 28029 Madrid, Spain.

Dr. Guallar: Johns Hopkins Bloomberg School of Public Health, 2024 East Monument Street, Room 2-639, Baltimore, MD 21205.

Author Contributions: Conception and design: M.P. Rayman, S. Stranges, B.A. Griffin, E. Guallar.

Analysis and interpretation of the data: M.P. Rayman, S. Stranges, B.A. Griffin, R. Pastor-Barriuso, E. Guallar.

Drafting of the article: M.P. Rayman, B.A. Griffin, R. Pastor-Barriuso.

Critical revision of the article for important intellectual content: M.P. Rayman, S. Stranges, B.A. Griffin, R. Pastor-Barriuso, E. Guallar.

Final approval of the article: M.P. Rayman, S. Stranges, B.A. Griffin, R. Pastor-Barriuso, E. Guallar.

Provision of study materials or patients: M.P. Rayman.

Statistical expertise: R. Pastor-Barriuso, E. Guallar.

Obtaining of funding: M.P. Rayman.

Administrative, technical, or logistic support: M.P. Rayman.

Collection and assembly of data: E. Guallar.


Ann Intern Med. 2011;154(10):656-665. doi:10.7326/0003-4819-154-10-201105170-00005
Text Size: A A A

Background: High selenium status has been linked to elevated blood cholesterol levels in cross-sectional studies.

Objective: To investigate the effect of selenium supplementation on plasma lipids.

Design: Randomized, placebo-controlled, parallel-group study stratified by age and sex. Participants, research nurses, and persons assessing outcomes were blinded to treatment assignment. (International Standard Randomised Controlled Trial Number Register registration number: ISRCTN25193534)

Setting: 4 general practices in the United Kingdom.

Participants: 501 volunteers aged 60 to 74 years.

Intervention: Participants received selenium, 100 mcg/d (n = 127), 200 mcg/d (n = 127), or 300 mcg/d (n = 126), as high-selenium yeast or a yeast-based placebo (n = 121) for 6 months.

Measurements: Total and high-density lipoprotein (HDL) cholesterol concentrations were measured in nonfasting plasma samples stored from participants in the UK PRECISE (United Kingdom PREvention of Cancer by Intervention with SElenium) Pilot Study at baseline (n = 454) and at 6 months (n = 394). Non-HDL cholesterol levels were calculated.

Results: Mean plasma selenium concentration was 88.8 ng/g (SD, 19.2) at baseline and increased statistically significantly in the treatment groups. The adjusted difference in change in total cholesterol levels for selenium compared with placebo was −0.22 mmol/L (−8.5 mg/dL) (95% CI, −0.42 to −0.03 mmol/L [−16.2 to −1.2 mg/dL]; P = 0.02) for 100 mcg of selenium per day, −0.25 mmol/L (−9.7 mg/dL) (CI, −0.44 to −0.07 mmol/L [−17.0 to −2.7 mg/dL]; P = 0.008) for 200 mcg of selenium per day, and −0.07 mmol/L (−2.7 mg/dL) (CI, −0.26 to 0.12 mmol/L [−10.1 to 4.6 mg/dL]; P = 0.46) for 300 mcg of selenium per day. Similar reductions were observed for non-HDL cholesterol levels. There was no apparent difference in change in HDL cholesterol levels with 100 and 200 mcg of selenium per day, but the difference was an adjusted 0.06 mmol/L (2.3 mg/dL) (CI, 0.00 to 0.11 mmol/L [0.0 to 4.3 mg/dL]; P = 0.045) with 300 mcg of selenium per day. The total–HDL cholesterol ratio decreased progressively with increasing selenium dose (overall P = 0.01).

Limitation: The duration of supplementation was limited, as was the age range of the participants.

Conclusion: Selenium supplementation seemed to have modestly beneficial effects on plasma lipid levels in this sample of persons with relatively low selenium status. The clinical significance of the findings is unclear and should not be used to justify the use of selenium supplementation as additional or alternative therapy for dyslipidemia. This is particularly true for persons with higher selenium status, given the limitations of the trial and the potential additional risk in other metabolic dimensions.

Primary Funding Source: The Cancer Research Campaign (now Cancer Research UK) and the University of Surrey.

Editors' Notes
Context

  • Observational studies of selenium status offer conflicting evidence that deficiency is associated with heart disease, whereas nondeficiency is associated with dyslipidemia.

Contribution

  • In this randomized trial, selenium supplementation was associated with modest reductions in total and non–high-density lipoprotein cholesterol levels. At higher doses, selenium supplementation was associated with increases in high-density lipoprotein cholesterol levels.

Caution

  • The trial lasted only 6 months and enrolled only older patients.

Implication

  • Selenium supplementation modestly improves lipid profiles. The findings more definitively identify the effects of selenium on lipids because they are trial-based. However, the findings do not justify the use of selenium for dyslipidemia because the clinical significance of the changes is unknown and selenium supplementation may have other side effects.

—The Editors

Since 1982, when Salonen and colleagues (1) showed a significant 2- to 3-fold increase in cardiovascular morbidity and mortality for persons with low serum selenium concentrations, there has been interest in the link between selenium status and cardiovascular disease. In theory, potential cardiovascular benefits are supported by the ability of selenoproteins, such as glutathione peroxidase and selenoprotein S, to combat the oxidative modification of lipids, inhibit platelet aggregation, and reduce inflammation (29). Despite a solid biological rationale for a beneficial effect of optimal selenoprotein activity and concentration on cardiovascular health, evidence that selenium status affects coronary heart disease risk is equivocal (1014). In contrast to the putative benefits of selenium, several cross-sectional studies have shown an association between higher selenium status and elevated plasma or serum concentrations of total or low-density lipoprotein cholesterol (1523). However, some studies have also linked higher selenium status to higher high-density lipoprotein (HDL) cholesterol levels (1920, 2324). Although no effect of selenium supplementation on serum cholesterol level was found in 2 small trials (2526) because of their low power, we cannot conclude that the cross-sectional association observed between selenium and serum lipid levels is not driven by selenium.

The UK PRECISE (United Kingdom PREvention of Cancer by Intervention with SElenium) Pilot Study, in which 501 elderly volunteers were randomly assigned to receive 6-month treatment with 100, 200, or 300 mcg of selenium per day as high-selenium yeast or a placebo yeast (2728), provided a unique opportunity to investigate the effect of selenium supplementation on plasma lipids in a considerably larger group than has been studied previously. Based on previous observational evidence, our working hypothesis was that selenium supplementation would increase total plasma cholesterol levels, thus providing supporting evidence for the cross-sectional association. A secondary aim was to determine the relative contributions of HDL and non-HDL cholesterol levels to this effect.

Design

The UK PRECISE Pilot Study for the planned international PRECISE trial was designed to assess the viability of conducting the trial in the United Kingdom. It was a double-blind, placebo-controlled, 4-group (allocation ratio was 1:1:1:1), parallel-group study stratified by age and sex and conducted at 4 sites in the United Kingdom (International Standard Randomised Controlled Trial Number Register registration number: ISRCTN25193534), as described elsewhere (2728). Funding was never secured for the international PRECISE trial, which did not therefore take place.

Sample

No formal power calculations were done for the pilot study, and the sample size (n = 501) was chosen to give a sufficient number to be able to draw reasonable inferences about recruitment, adherence, and loss to follow-up while keeping the cost within reasonable bounds.

Setting and Participants

Volunteers were recruited from 4 general practices in different parts of the United Kingdom (Table 1) affiliated with the Medical Research Council General Practice Research Framework. The local research ethics committees (South Tees, Worcestershire Health Authority, Norwich District, and Great Yarmouth and Waveney [under reciprocal arrangements with the Norwich District local research ethics committee]) approved the study, and participants provided written informed consent.

Table Jump PlaceholderTable 1.  Descriptive Baseline Characteristics, Overall and by Group
Randomization and Interventions

From June 2000 to July 2001, research nurses recruited similar numbers of men and women from each of 3 age groups: 60 to 64 years, 65 to 69 years, and 70 to 74 years. Exclusion criteria were a Southwest Oncology Group performance status score greater than 1 (that is, incapable of carrying out light housework or office work); active liver or kidney disease; a previous diagnosis of cancer (excluding nonmelanoma skin cancer); diagnosed HIV infection; receipt of immunosuppressive therapy at recruitment; diminished mental capacity; or receipt of selenium supplements, 50 mcg/d, in the previous 6 months (by patient report). Written informed consent was obtained from all participants after they had been given a detailed description of the study, time to consider, and the opportunity to ask questions.

Computer-generated, random, permuted blocks stratified by general practice, sex, and age group were used to generate a randomized list at the independent randomization service at the Clinical Trials and Statistics Unit, Institute of Cancer Research (Sutton, United Kingdom). The Medical Research Council General Practice Research Framework stockroom supplied the trial tablets to the practices according to this list. The research nurses in the practices telephoned the randomization service at the Clinical Trials and Statistics Unit, which had a dedicated telephone line, to obtain an anonymous code (that is, trial number) for each eligible volunteer. The volunteer was then given the precoded tablets bearing that trial number.

After a 4-week, placebo run-in phase, 501 volunteers were randomly assigned in equal numbers to receive 1 of 4 treatment regimens: placebo or 100, 200, or 300 mcg of selenium per day for a minimum of 6 months. The intervention agent was the high-selenium yeast SelenoPrecise (Pharma Nord, Vejle, Denmark) or an identical placebo yeast (comprising 250 mg of yeast placebo, 80 mg of cellulose, 65 mg of dicalcium phosphate, and ≤5 mg of other inactive ingredients). Demographic data; medical history; and other health-related information, including medication and supplement use, were collected at baseline. Participants provided a nonfasting blood sample at both baseline and 6 months.

Blinding and Outcomes

Participants, research nurses, other general practice personnel, investigators, and persons who analyzed the data were blinded to treatment. The defined prespecified primary outcome was satisfactory recruitment, retention, and adherence of the volunteers so that the viability of conducting the main PRECISE trial in the United Kingdom could be ascertained. Prespecified secondary outcome measures were the effect of selenium supplementation on mood and thyroid function (2728).

Nonspecified outcomes that were later examined on stored samples were based on literature reports that plasma total homocysteine concentration was inversely associated with plasma selenium concentration (29) and that plasma total cholesterol level was positively associated with plasma or serum selenium concentration. Thus, the effect of selenium supplementation on plasma total homocysteine and B vitamin concentrations were investigated, although only in the groups that received placebo and 100 and 300 mcg of selenium per day (for reasons of financial constraint) (29), and the effects of selenium supplementation on plasma total cholesterol and HDL cholesterol levels were measured in the current study.

Follow-up

The last planned 6-month follow-up visit (with blood draw) was in January 2002. However, volunteers, who were to be the first cohort of the main PRECISE trial, continued treatment and 6-month follow-up visits until mid-2003 when it became clear that the international study was not going to be funded.

Heparinized plasma was prepared and frozen at the practices. Plasma samples were transferred to the University of Surrey Guildford, United Kingdom, on dry ice, where they were stored at −80 °C. At the 6-month follow-up visit, questionnaires were used to see whether any new symptoms or illnesses occurred since randomization and use of medication and supplements had changed. Adherence was assessed by pill count—participants were considered adherent if they took at least 80% of their allocated tablets. In addition, each participant's plasma selenium concentration was compared with the mean concentration of the group to detect any outliers—nonadherent participants or “drop-ins” (nonprotocol use of over-the-counter selenium). Reasons for participant withdrawal were noted.

Selenium Measurement

Lithium–heparin plasma was analyzed for selenium at Central Science Laboratory, Sand Hutton, United Kingdom, by using hydride-generation inductively coupled-plasma mass spectrometry, as described elsewhere (28). Accuracy was assured by good performance on the analysis of certified reference materials (28).

Plasma Lipid Measurement

Total and HDL cholesterol levels were measured on the remaining stored plasma samples (454 participants at baseline and 394 participants at 6 months) by using routine colorimetric assays (Randox, Crumlin, United Kingdom) on an IL ILab 650 Clinical Chemistry Analyzer (Instrumentation Laboratory, Warrington, United Kingdom). Non-HDL cholesterol level was calculated as total cholesterol minus HDL cholesterol. The mean values for the quality-control total and HDL cholesterol samples were within the manufacturer's certified ranges. The intra- and interassay coefficients of variation were less than 1.4% for total cholesterol levels and less than 2.6% for HDL cholesterol levels.

Statistical Analysis

The effect of selenium supplementation on plasma levels of total, HDL, and non-HDL cholesterol and on the total–HDL cholesterol ratio was estimated by using linear mixed models for longitudinal data with random between-participant variations in baseline lipid levels (intercepts) and lipid changes over time (slopes). The models are specified in detail in the Appendix. Participants were randomly assigned to a group regardless of adherence (intention-to-treat assignment). The reported P values are 2-sided and were not adjusted for multiple testing. Analyses of covariance restricted to participants with lipid measurements available both at baseline and at 6 months (n = 374) resulted in virtually identical results (data not shown).

In addition to the intention-to-treat analysis, we conducted additional analyses of the cross-sectional association between baseline plasma selenium concentrations and plasma lipids and analyses of the longitudinal association between change in plasma selenium concentrations and change in lipid levels. The cross-sectional analyses do not reflect the trial intervention but enable the results to be compared with those of previous observational studies (1524). The longitudinal analyses reflect the trial intervention and are equivalent to per-protocol analyses by using the observed change in plasma selenium concentrations as the exposure metric. Cross-sectional and longitudinal associations were obtained from random-intercept and random-slope linear mixed models for longitudinal data. The models are specified in detail in the Appendix. Statistical analyses were done with Stata, version 11 (StataCorp, College Station, Texas).

Role of the Funding Source

The Cancer Research Campaign (now Cancer Research UK) funded the UK PRECISE Pilot Study. The Cardiovascular Disease, Diabetes & Metabolism Research Theme, University of Surrey, provided funding for the lipid measurements. The funding sources had no role in the study design; data collection, analysis, or interpretation; or decision to submit the manuscript for publication.

Participants

Of 501 randomly assigned participants, 34 withdrew from treatment (Figure). There was no statistically significant difference in numbers of participants who withdrew across treatment groups (P = 0.31). Our report is restricted to the 474 participants (94.6%) who had at least 1 lipid measurement available either at baseline or at 6 months (lipid measurements were available for 454 participants at baseline, 394 at 6 months, and 374 both at baseline and at 6 months) (Figure). Characteristics of participants with and without available lipid measurements did not differ (data not shown). A total of 94% of the 474 participants missed less than 10% of the total number of study tablets, according to pill count. Nonprotocol use of over-the-counter selenium (drop-ins) was assessed by inspection of the histogram of plasma selenium concentration in the placebo group at 6 months. Four of 107 participants who received placebo (3.7%) had a selenium status more than 2 SDs above the mean, which is reasonably consistent with the 2.5% expected by chance (being approximately normally distributed). We have assumed, therefore, that drop-ins were rare. Three participants reported receiving supplements containing 25 mcg of selenium. Because this was within the daily limit of 50 mcg imposed by our exclusion criteria, the participants have been included in the analysis. Twenty-two participants were receiving lipid-lowering medications, with no statistically significant difference across the groups (Table 1).

Intention-to-Treat Analysis

Overall mean plasma selenium concentration at baseline was 88.8 ng/g (SD, 19.2) (equivalent to 91.2 mcg/L [SD, 19.7]) (30). There were no statistically significant differences between treatment groups at baseline in plasma selenium concentration (P = 0.83) or in other participant characteristics, including lipid concentrations (Table 1). After 6 months of supplementation, plasma selenium had increased statistically significantly in the treatment groups but was unchanged in the placebo group (Table 2). In addition, the 3 selenium groups had statistically significantly different lipid concentrations than those in the placebo group for all lipid variables (Table 2).

Table Jump PlaceholderTable 2.  Effect of Selenium Supplementation on Lipid and Plasma Selenium Concentrations After 6 Months

Selenium supplementation of 100 and 200 mcg per day for 6 months decreased total and non-HDL cholesterol levels. There was no statistically significant effect of the 300-mcg daily dose on total or non-HDL cholesterol levels, but this dose statistically significantly increased HDL cholesterol levels. The total–HDL cholesterol ratio was statistically significantly reduced after treatment with the 200- and 300-mcg daily selenium doses. Exclusion of the 22 participants who received lipid-lowering medication did not materially alter the results. In subgroup analyses, there were no statistically significant differences across study centers (Appendix Table) or by sex, category of body mass index, baseline plasma selenium concentration, or baseline lipid concentrations (data not shown).

Table Jump PlaceholderAppendix Table.  Effect of Selenium Supplementation on Changes in Lipid Concentrations After 6 Months, by Study Center
Association of Plasma Selenium Concentration With Lipid Levels

In cross-sectional analyses at baseline, increasing plasma selenium was associated with increasing total and HDL cholesterol levels and with decreasing total–HDL cholesterol ratio (Table 3 and Appendix Figure 1). In longitudinal analyses, increasing plasma selenium concentrations from baseline to 6 months were associated with decreasing total cholesterol levels, non-HDL cholesterol levels, and the ratio of total–HDL cholesterol and with increasing HDL cholesterol levels (Table 4 and Appendix Figure 2).

Table Jump PlaceholderTable 3.  Cross-sectional Association of Plasma Selenium Concentrations With Lipid Concentrations at Baseline
Grahic Jump Location
Appendix Figure 1.
Lipid concentrations, by plasma selenium concentration at baseline.

Curves represent mean baseline lipid concentrations (solid line) and their 95% CIs (dashed line) based on restricted quadratic splines for baseline selenium concentrations, with knots at the 5th, 50th, and 95th percentiles (61.0, 87.8, and 119.0 ng/g, respectively). Results were obtained from linear mixed models with random between-participant variations in baseline lipid levels and adjusted for age, sex, center, smoking status, body mass index, and use of lipid-lowering medications (Appendix). To convert cholesterol values to mg/dL, divide by 0.02586. Scatterplots represent adjusted baseline concentrations of selenium and lipids. HDL = high-density lipoprotein.

Grahic Jump Location
Table Jump PlaceholderTable 4.  Longitudinal Association of Changes in Plasma Selenium Concentration With Changes in Lipid Concentrations After 6 Months
Grahic Jump Location
Appendix Figure 2.
Changes in lipid concentrations after 6 months, by change in plasma selenium concentration.

Curves represent mean changes in lipid concentrations from baseline to 6 months (solid line) and their 95% CIs (dashed line) based on restricted quadratic splines for within-participant changes in selenium concentration, with knots at the 5th, 50th, and 95th percentiles (−8.0, 74.0, and 173.9 ng/g, respectively). Results were obtained from linear mixed models with random between-participant variations in both baseline lipid levels and lipid changes over time and adjusted for baseline selenium concentration, age, sex, center, smoking status, body mass index, and use of lipid-lowering medications (Appendix). Scatterplots represent adjusted changes in selenium and lipid concentrations over time for participants randomly assigned to receive placebo (gray dots) or 100 mcg (blue dots), 200 mcg (green dots), or 300 mcg (red dots) of selenium per day. To convert cholesterol values to mg/dL, divide by 0.02586. HDL = high-density lipoprotein.

Grahic Jump Location
Adverse Events

No serious adverse events occurred. Twelve adverse events were reported, which were principally stomach or abdominal discomfort. These were equally associated with selenium or placebo and were not dependent on dose (data not shown).

Contrary to our original hypothesis, selenium supplementation of 100 and 200 mcg per day statistically significantly decreased total and non-HDL cholesterol concentrations. Supplementation of 300 mcg per day had a smaller and nonstatistically significant effect on total and non-HDL cholesterol levels, but it was the only dose that statistically significantly increased HDL cholesterol levels. The total–HDL cholesterol ratio decreased progressively with increasing selenium dose. Two previous trials of the effect of selenium supplementation on plasma lipids in 27 participants (25) and 40 participants (26) did not show any effects, although 1 trial reported a statistically significant decrease in the total–HDL cholesterol ratio after excluding 1 participant with a total cholesterol level less than 3.5 mmol/L (<135.1 mg/dL) (25).

The effect of selenium on lipid levels has also been extensively studied in experimental animal models in which selenium supplementation decreased plasma total and low-density lipoprotein cholesterol levels and increased HDL cholesterol levels, whereas selenium deficiency had the opposite effect (3134). However, the relevance of such studies to humans is questionable.

Although we do not know the mechanisms underlying the effects of selenium on plasma lipids, several pathways involving selenium or selenoproteins are known to interact with both lipids and lipoproteins. Selenium is incorporated into selenoproteins as selenocysteine (Sec), which is dependent on its specific transfer RNA, designated as tRNA[Ser]Sec. Preventing selenoprotein synthesis by targeted removal of the tRNA[Ser]Sec gene (Trsp) in mouse hepatocytes elevated plasma cholesterol concentration, increased the amount of apolipoprotein E, enhanced the expression of genes involved in cholesterol biosynthesis, and decreased the expression of those involved in cholesterol metabolism and transport (35). Rescuing the mice with mutant Trsp transgenes that restored or partially restored several selenoprotein messenger RNAs and returned the levels of apolipoprotein E and cholesterol to those in normal mice suggests that 1 or more of these selenoproteins can decrease apolipoprotein E and plasma cholesterol concentrations (35). Furthermore, selenoprotein and cholesterol synthesis are connected through the common mevalonate pathway of isoprenoid biosynthesis (36). The rate-limiting enzyme in the mevalonate pathway, 3-hydroxy-3-methylglutaryl coenzyme A reductase, is inhibited by statins resulting in a decrease in plasma cholesterol levels. Of interest, patients with dyslipidemia who received treatment with statins had lower plasma selenium concentrations than patients who did not receive treatment (37). Finally, selenium supplementation increases the production of 15-deoxy-Δ-12,14-prostaglandin J2 (7, 3839), a known peroxisome proliferator-activated receptor-γ ligand. Activation of peroxisome proliferator-activated receptor-γ can reduce the concentration of sterol regulatory element-binding protein-2, resulting in a reduction of cholesterol synthesis (40). Our cross-sectional analysis of baseline data is consistent with findings from most observational studies that showed associations between high selenium status and elevated concentrations of total and HDL cholesterol (1524), although the association with HDL cholesterol was particularly strong in our study. The results from our randomized, controlled trial, however, indicate that selenium supplementation may be more likely to reduce total cholesterol levels and increase HDL cholesterol levels. As a consequence, it now seems more likely that 1 or more unmeasured factors may increase both plasma selenium and total and non-HDL cholesterol levels, thus explaining the consistent cross-sectional associations.

The implications of these findings for cardiovascular prevention are unclear. Randomized trials of selenium-containing supplements have not demonstrated a protective effect on cardiovascular disease or mortality end points (1013). A meta-analysis of 14 cohort studies found a modest inverse association between biomarkers of selenium status, such as blood or toenail selenium concentrations, and the risk for coronary heart disease (10). However, no associations were seen between toenail selenium concentration and measures of subclinical atherosclerosis—that is, carotid intima–media thickness and coronary artery calcium score—in a study of 3112 young U.S. adults (41).

We think that the association of selenium status with cardiovascular outcomes will probably be influenced by the level of selenium status of the population studied. For instance, the absence of an effect of selenium supplementation on cardiovascular end points reported in 2 studies (11, 13) that used data from the NPC (Nutritional Prevention of Cancer) trial (mean plasma selenium concentration, 114 mcg/L) and SELECT (Selenium and Vitamin E Cancer Prevention Trial) (mean serum selenium concentration, 136 mcg/L) were in samples in which most selenoproteins, including glutathione peroxidase, would already have been optimized at baseline (42). Hence, if the potential beneficial effect of selenium on plasma lipids is dependent on increasing selenoprotein concentrations, no effect of supplementation would have been seen. By contrast, the risk for a cardiovascular event in patients with suspected coronary artery disease and considerably lower selenium status (mean plasma selenium, 74 mcg/L) followed for a median period of 4.7 years was shown to be inversely related to baseline intracellular glutathione peroxidase activity, illustrating a potentially beneficial effect of selenoprotein activity on cardiovascular risk (6). Observational findings from the NHANES III (Third National Health and Nutrition Examination Survey) also support the idea that baseline selenium status may influence the apparent effect of supplementation insofar as they showed that above a certain plasma selenium concentration (at which we may assume that relevant selenoproteins have been optimized), there was no further advantage of higher selenium status in reducing cardiovascular mortality (43).

This study has several limitations. We have shown an effect only on plasma lipids and not on cardiovascular disease risk, with lipid concentrations being, at best, only a surrogate measure of risk. Blood samples were collected in the nonfasting state, so total and HDL cholesterol concentration are likely to be accurate (44), but we could not measure plasma triglyceride levels, an additional coronary heart disease risk factor. Follow-up was only for 6 months, and we do not know whether the effect of selenium supplementation would differ over a longer period. Finally, with our sample size, we did not have sufficient power to evaluate subgroup effects reliably. In view of the relatively low baseline selenium status of the population, the limited duration of the intervention (6 months) and the limited age range of the group (60 to 74 years), further clinical trials are needed to corroborate our findings.

Our study was done in a sample in which few participants had maximized selenoprotein activities or concentrations. This would not have been the case in the United States, where mean serum selenium concentration, recently measured as 136.7 mcg/L (SD, 18.9) (45), is sufficient to optimize selenoprotein status in all but a few participants. In view of the potential public health implications of hyperlipidemia and the widespread use of selenium-containing supplements, our results are reassuring. However, the full range of cardiometabolic effects of selenium and its potential role in cardiovascular disease risk need to be further explored in larger randomized trials in populations that span the worldwide range of selenium intake.

In summary, selenium supplementation seemed to have modestly beneficial effects on plasma lipid levels in this sample of relatively low selenium status. Although the findings were statistically significant, their clinical significance is less clear. In that context, the findings perhaps more importantly suggest absence of harm of selenium supplementation in this set of metabolic measures (in which past evidence has suggested possible adverse effects) and should not be used to justify the use of selenium supplementation as additional or alternative therapy for dyslipidemia, particularly in persons with higher selenium status, given the limitations of the trial and the potential additional risk in other metabolic dimensions (46).

Salonen JT, Alfthan G, Huttunen JK, Pikkarainen J, Puska P.  Association between cardiovascular death and myocardial infarction and serum selenium in a matched-pair longitudinal study. Lancet. 1982; 2:175-9.
PubMed
CrossRef
 
Nève J.  Selenium as a risk factor for cardiovascular diseases. J Cardiovasc Risk. 1996; 3:42-7.
PubMed
 
Brigelius-Flohé R, Banning A, Schnurr K.  Selenium-dependent enzymes in endothelial cell function. Antioxid Redox Signal. 2003; 5:205-15.
PubMed
 
Sattler W, Maiorino M, Stocker R.  Reduction of HDL- and LDL-associated cholesterylester and phospholipid hydroperoxides by phospholipid hydroperoxide glutathione peroxidase and Ebselen (PZ 51). Arch Biochem Biophys. 1994; 309:214-21.
PubMed
 
Salonen JT, Salonen R, Seppänen K, Kantola M, Parviainen M, Alfthan G. et al.  Relationship of serum selenium and antioxidants to plasma lipoproteins, platelet aggregability and prevalent ischaemic heart disease in Eastern Finnish men. Atherosclerosis. 1988; 70:155-60.
PubMed
 
Blankenberg S, Rupprecht HJ, Bickel C, Torzewski M, Hafner G, Tiret L, et al. AtheroGene Investigators.  Glutathione peroxidase 1 activity and cardiovascular events in patients with coronary artery disease. N Engl J Med. 2003; 349:1605-13.
PubMed
 
Vunta H, Davis F, Palempalli UD, Bhat D, Arner RJ, Thompson JT. et al.  The anti-inflammatory effects of selenium are mediated through 15-deoxy-Delta12,14-prostaglandin J2 in macrophages. J Biol Chem. 2007; 282:17964-73.
PubMed
 
Gao Y, Hannan NR, Wanyonyi S, Konstantopolous N, Pagnon J, Feng HC. et al.  Activation of the selenoprotein SEPS1 gene expression by pro-inflammatory cytokines in HepG2 cells. Cytokine. 2006; 33:246-51.
PubMed
 
Curran JE, Jowett JB, Elliott KS, Gao Y, Gluschenko K, Wang J. et al.  Genetic variation in selenoprotein S influences inflammatory response. Nat Genet. 2005; 37:1234-41.
PubMed
 
Flores-Mateo G, Navas-Acien A, Pastor-Barriuso R, Guallar E.  Selenium and coronary heart disease: a meta-analysis. Am J Clin Nutr. 2006; 84:762-73.
PubMed
 
Stranges S, Marshall JR, Trevisan M, Natarajan R, Donahue RP, Combs GF. et al.  Effects of selenium supplementation on cardiovascular disease incidence and mortality: secondary analyses in a randomized clinical trial. Am J Epidemiol. 2006; 163:694-9.
PubMed
 
Bjelakovic G, Nikolova D, Gluud LL, Simonetti RG, Gluud C.  Mortality in randomized trials of antioxidant supplements for primary and secondary prevention: systematic review and meta-analysis. JAMA. 2007; 297:842-57.
PubMed
 
Lippman SM, Klein EA, Goodman PJ, Lucia MS, Thompson IM, Ford LG. et al.  Effect of selenium and vitamin E on risk of prostate cancer and other cancers: the Selenium and Vitamin E Cancer Prevention Trial (SELECT). JAMA. 2009; 301:39-51.
PubMed
 
Stranges S, Navas-Acien A, Rayman MP, Guallar E.  Selenium status and cardiometabolic health: state of the evidence. Nutr Metab Cardiovasc Dis. 2010; 20:754-60.
PubMed
 
Jossa F, Trevisan M, Krogh V, Farinaro E, Giumetti D, Fusco G. et al.  Serum selenium and coronary heart disease risk factors in southern Italian men. Atherosclerosis. 1991; 87:129-34.
PubMed
 
Suadicani P, Hein HO, Gyntelberg F.  Serum selenium concentration and risk of ischaemic heart disease in a prospective cohort study of 3000 males. Atherosclerosis. 1992; 96:33-42.
PubMed
 
Gámez C, Ruiz-López D, Artacho R, Navarro M, Puerta A, López C.  Serum selenium in institutionalized elderly subjects and relation to other nutritional markers [Letter]. Clin Chem. 1997; 43:693-4.
PubMed
 
Coudray C, Roussel AM, Mainard F, Arnaud J, Favier A.  Lipid peroxidation level and antioxidant micronutrient status in a pre-aging population; correlation with chronic disease prevalence in a French epidemiological study (Nantes, France). J Am Coll Nutr. 1997; 16:584-91.
PubMed
 
Bleys J, Navas-Acien A, Stranges S, Menke A, Miller ER 3rd, Guallar E.  Serum selenium and serum lipids in US adults. Am J Clin Nutr. 2008; 88:416-23.
PubMed
 
Laclaustra M, Stranges S, Navas-Acien A, Ordovas JM, Guallar E.  Serum selenium and serum lipids in US adults: National Health and Nutrition Examination Survey (NHANES) 2003-2004. Atherosclerosis. 2010; 210:643-8.
PubMed
 
Stranges S, Laclaustra M, Ji C, Cappuccio FP, Navas-Acien A, Ordovas JM. et al.  Higher selenium status is associated with adverse blood lipid profile in British adults. J Nutr. 2010; 140:81-7.
PubMed
 
Yang KC, Lee LT, Lee YS, Huang HY, Chen CY, Huang KC.  Serum selenium concentration is associated with metabolic factors in the elderly: a cross-sectional study. Nutr Metab (Lond). 2010; 7:38.
PubMed
 
Bates CJ, Thane CW, Prentice A, Delves HT.  Selenium status and its correlates in a British national diet and nutrition survey: people aged 65 years and over. J Trace Elem Med Biol. 2002; 16:1-8.
PubMed
 
Salonen JT, Salonen R, Seppänen K, Kantola M, Parviainen M, Alfthan G. et al.  Relationship of serum selenium and antioxidants to plasma lipoproteins, platelet aggregability and prevalent ischaemic heart disease in Eastern Finnish men. Atherosclerosis. 1988; 70:155-60.
PubMed
 
Luoma PV, Sotaniemi EA, Korpela H, Kumpulainen J.  Serum selenium, glutathione peroxidase activity and high-density lipoprotein cholesterol—effect of selenium supplementation. Res Commun Chem Pathol Pharmacol. 1984; 46:469-72.
PubMed
 
Yu SY, Mao BL, Xiao P, Yu WP, Wang YL, Huang CZ. et al.  Intervention trial with selenium for the prevention of lung cancer among tin miners in Yunnan, China. A pilot study. Biol Trace Elem Res. 1990; 24:105-8.
PubMed
 
Rayman M, Thompson A, Warren-Perry M, Galassini R, Catterick J, Hall E. et al.  Impact of selenium on mood and quality of life: a randomized, controlled trial. Biol Psychiatry. 2006; 59:147-54.
PubMed
 
Rayman MP, Thompson AJ, Bekaert B, Catterick J, Galassini R, Hall E. et al.  Randomized controlled trial of the effect of selenium supplementation on thyroid function in the elderly in the United Kingdom. Am J Clin Nutr. 2008; 87:370-8.
PubMed
 
Bekaert B, Cooper ML, Green FR, McNulty H, Pentieva K, Scott JM. et al.  Effect of selenium status and supplementation with high-selenium yeast on plasma homocysteine and B vitamin concentrations in the UK elderly. Mol Nutr Food Res. 2008; 52:1324-33.
PubMed
 
Lide DR, ed.  CRC Handbook of Chemistry and Physics. 74th ed. Boca Raton, FL: CRC Pr; 1993-1994.
 
Wolf NM, Mueller K, Hirche F, Most E, Pallauf J, Mueller AS.  Study of molecular targets influencing homocysteine and cholesterol metabolism in growing rats by manipulation of dietary selenium and methionine concentrations. Br J Nutr. 2010; 104:520-32.
PubMed
 
Dhingra S, Bansal MP.  Hypercholesterolemia and apolipoprotein B expression: regulation by selenium status. Lipids Health Dis. 2005; 4:28.
PubMed
 
Mazur A, Nassir F, Gueux E, Moundras C, Bellanger J, Grolier P. et al.  Diets deficient in selenium and vitamin E affect plasma lipoprotein and apolipoprotein concentrations in the rat. Br J Nutr. 1996; 76:899-907.
PubMed
 
Wójcicki J, Rózewicka L, Barcew-Wiszniewska B, Samochowiec L, Juźwiak S, Kadłubowska D. et al.  Effect of selenium and vitamin E on the development of experimental atherosclerosis in rabbits. Atherosclerosis. 1991; 87:9-16.
PubMed
 
Sengupta A, Carlson BA, Hoffmann VJ, Gladyshev VN, Hatfield DL.  Loss of housekeeping selenoprotein expression in mouse liver modulates lipoprotein metabolism. Biochem Biophys Res Commun. 2008; 365:446-52.
PubMed
 
Moosmann B, Behl C.  Selenoprotein synthesis and side-effects of statins. Lancet. 2004; 363:892-4.
PubMed
 
Arnaud J, Akbaraly TN, Hininger-Favier I, Berr C, Roussel AM.  Fibrates but not statins increase plasma selenium in dyslipidemic aged patients—the EVA study. J Trace Elem Med Biol. 2009; 23:21-8.
PubMed
 
Ricote M, Li AC, Willson TM, Kelly CJ, Glass CK.  The peroxisome proliferator-activated receptor-gamma is a negative regulator of macrophage activation. Nature. 1998; 391:79-82.
PubMed
 
Touyz RM, Schiffrin EL.  Peroxisome proliferator-activated receptors in vascular biology-molecular mechanisms and clinical implications. Vascul Pharmacol. 2006; 45:19-28.
PubMed
 
Klopotek A, Hirche F, Eder K.  PPAR gamma ligand troglitazone lowers cholesterol synthesis in HepG2 and Caco-2 cells via a reduced concentration of nuclear SREBP-2. Exp Biol Med (Maywood). 2006; 231:1365-72.
PubMed
 
Xun P, Liu K, Morris JS, Daviglus ML, He K.  Longitudinal association between toenail selenium levels and measures of subclinical atherosclerosis: the CARDIA trace element study. Atherosclerosis. 2010; 210:662-7.
PubMed
 
Duffield AJ, Thomson CD, Hill KE, Williams S.  An estimation of selenium requirements for New Zealanders. Am J Clin Nutr. 1999; 70:896-903.
PubMed
 
Bleys J, Navas-Acien A, Guallar E.  Serum selenium levels and all-cause, cancer, and cardiovascular mortality among US adults. Arch Intern Med. 2008; 168:404-10.
PubMed
 
Halloran P, Roetering H, Pisani T, van den Berg B, Cobbaert C.  Reference standardization and analytical performance of a liquid homogeneous high-density lipoprotein cholesterol method compared with chemical precipitation method. Arch Pathol Lab Med. 1999; 123:317-26.
PubMed
 
Centers for Disease Control and Prevention (CDC); National Center for Health Statistics (NCHS).  National Health and Nutrition Examination Survey Data. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2003-2004. Accessed atwww.cdc.gov/nchs/nhanes/nhanes2003-2004/nhanes03_04.htmon 23 March 2011.
 
Stranges S, Marshall JR, Natarajan R, Donahue RP, Trevisan M, Combs GF. et al.  Effects of long-term selenium supplementation on the incidence of type 2 diabetes: a randomized trial. Ann Intern Med. 2007; 147:217-23.
PubMed
 
Raudenbush SW, Bryk AS.  Hierarchical Linear Models: Applications and Data Analysis Methods. 2nd ed. Thousand Oaks, CA: Sage; 2002.
 
Albert PS.  Longitudinal data analysis (repeated measures) in clinical trials. Stat Med. 1999; 18:1707-32.
PubMed
 
Greenland S.  Dose-response and trend analysis in epidemiology: alternatives to categorical analysis. Epidemiology. 1995; 6:356-65.
PubMed
 
Diggle PJ, Heagerty P, Liang KY, Zeger SL.  Analysis of Longitudinal Data. 2nd ed. Oxford, United Kingdom: Oxford Univ Pr; 2002.
 
Appendix: Statistical Appendix

Note: Some formatting cannot be preserved in the HTML version of this Appendix. For accurate formatting, the reader is referred to the PDF.

Efficacy of Selenium Supplementation, by Intention-to-Treat Analysis

To evaluate the effect of selenium supplementation on plasma levels of total cholesterol, HDL cholesterol, non-HDL cholesterol, and total–HDL cholesterol ratio, we developed linear mixed models for longitudinal data by using a 2-level hierarchical approach (47). At the first (within-participant) level, the change in plasma lipid levels Yti from baseline t = 0 to 6-month visit t = 1 for participant i = 1,…, 474 was described through the linear relation

in which α0i and α1i were the expected lipid levels at baseline and the expected change in lipid levels from baseline to 6 months, respectively, for that participant, and the within-participant errors εti were assumed to be independent and normally distributed with mean 0 and constant variance σ2.

The second level represented the variability in the parameters α0i and α1i across participants. The individual baseline lipid levels α0i were allowed to vary by study center (Bungay, Guisborough, Bromsgrove, or Linthorpe) and randomized treatment assignment by including center indicators c1i, c2i, and c3i for all but 1 reference center and treatment indicators T1i, T2i, and T3i for the 3 active treatment groups (100, 200, and 300 mcg of selenium supplementation per day, respectively). The individual changes in lipid levels from baseline to 6 months α1i were also allowed to differ across treatment groups. Specifically, the second-level (between-participant) model was

in which the between-participant errors b0i and b1i were assumed to be independent and normally distributed with mean 0 and respective variances τ02 and τ12. Combining the 2 nested models, we obtained the linear mixed model

closely related to other mixed-effects models proposed for the analysis of longitudinal data in clinical trials (48).

In model (A1), the fixed effect β00 represented the mean baseline lipid level for the placebo group in the reference study center; the fixed effects β01, β02, and β03 represented the treatment-adjusted differences in mean baseline lipid levels for each study center compared with the reference center; the fixed effects β04, β05, and β06 represented the center-adjusted differences in mean baseline lipid levels for the 3 active treatment groups compared with placebo; and the random effect b0i represented the unexplained between-participant variation in baseline lipid levels.

Similarly, the fixed effect β10 represented the mean change in lipid levels from baseline to 6 months for the placebo group; the fixed effects β11, β12, and β13 represented the center-adjusted differences in mean longitudinal lipid changes for the 3 active treatment groups compared with placebo; and the random effect b1i represented the unexplained variation in longitudinal lipid changes across participants within the same treatment group.

The longitudinal effects of 100, 200, and 300 mcg of selenium supplementation per day on changes in levels of total cholesterol, HDL cholesterol, non-HDL cholesterol, and total–HDL cholesterol ratio, as estimated from coefficients β11, β12, and β13 in model (A1), are shown in Table 2. In a sensitivity analysis, we fitted standard analysis of covariance models relating plasma lipid concentrations at 6 months to treatment assignment, adjusting for baseline lipid levels and study center, with virtually identical findings (data not shown).

To evaluate potential heterogeneity of treatment effects across study centers, we included all 2- and 3-way interactions among time, treatment group, and center as fixed effects in model (A1). The Wald test P values for the joint null hypothesis that all 3-way interaction coefficients were simultaneously 0 were 0.07, 0.30, 0.12, and 0.80 for levels of total cholesterol, HDL cholesterol, non-HDL cholesterol, and total–HDL cholesterol ratio, respectively. The estimates of the effects of selenium supplementation on plasma lipid levels within each center were imprecise but consistent with the overall results (Appendix Table).

Cross-sectional and Longitudinal Associations Between Plasma Selenium and Lipid Concentrations

To assess the cross-sectional association of baseline selenium and lipid levels, as well as the longitudinal association between changes in plasma selenium and lipid levels over time, we developed additional linear mixed models for longitudinal data by using a 2-level hierarchical approach. The first-level model, which describes changes in within-participant plasma lipid levels Yti from baseline t = 0 to 6 months t = 1, was identical to that of the intention-to-treat analysis:

At the second level, the individual baseline lipid levels α0i were allowed to vary as a linear function of the individual baseline selenium levels x0i and other participant's characteristics measured at baseline z0i, including age (continuous), sex, study center (Bungay, Guisborough, Bromsgrove, or Linthorpe), smoking status (never, former, or current), body mass index (continuous), and use of lipid-lowering medications. The individual changes in lipid levels from baseline to 6 months α1i were linearly related to the individual changes in selenium levels over time x1i − x0i. Thus, the second-level model was

in which the between-participant errors b0i and b1i were assumed to be independent and normally distributed with mean 0 and respective variances τ02 and τ12. Combining the 2 nested models and noting that (x1i − x0i)t = xti − x0i, we obtained the linear mixed model

In model (A2), the first line on the right-hand side represented the cross-sectional association of selenium levels and other covariates with plasma lipids at baseline. The fixed effect β00 represented the mean baseline lipid level at 0 values of selenium and other baseline covariates; the fixed effect β01 represented the covariate-adjusted mean change in baseline lipid levels per unit increase in baseline selenium levels; the fixed effects β02 represented the adjusted coefficients associated with the other baseline covariates; and the random effect b0i represented the unexplained between-participant variation in baseline lipid levels. The mean changes in baseline levels of total cholesterol, HDL cholesterol, non-HDL cholesterol, and total–HDL cholesterol ratio per 50-ng/g (approximately the 10th to the 90th percentile) increase in baseline selenium levels estimated from model (A2) are shown in Table 3. To assess nonlinear relationships, we modified the baseline portion of this mixed model to estimate the mean differences in baseline lipid levels for the 3 highest quartiles of baseline selenium levels compared with the lowest quartile (Table 3), as well as the mean baseline lipid changes on the basis of restricted quadratic splines (49) for baseline selenium levels with knots at the 5th, 50th, and 95th percentiles (61.0, 87.8, and 119.0 ng/g, respectively) (Appendix Figure 1). In sensitivity analyses, we did standard multivariable linear regressions of plasma lipid concentrations on selenium levels by using only cross-sectional data from the baseline visit, with similar results (data not shown).

The second line of the linear mixed model (A2) represented the longitudinal association of changes in within-participant selenium levels with changes in within-participant plasma lipid levels adjusted for baseline covariates (50). The fixed effect β10 represented the mean change in lipid levels from baseline to 6 months when selenium remained unchanged; the fixed effect β11 represented the covariate-adjusted rate of change in lipid levels over time per unit change in selenium levels; and the random effect b1i represented the unexplained between-participant variation in lipid changes over time. The mean changes in levels of total cholesterol, HDL cholesterol, non-HDL cholesterol, and total–HDL cholesterol ratio over time for each 50-ng/g increase in within-participant selenium levels (approximately the mean increase after 6 months of selenium supplementation, 100 mcg/d) and for each quartile of change in selenium levels are shown in Table 4. To allow for nonlinear longitudinal effects, we replaced the linear term xti − x0i in model (A2) with a restricted quadratic spline function (49) for within-participant selenium changes with knots at the 5th, 50th, and 95th percentiles (−8.0, 74.0, and 173.9 ng/g, respectively). The resulting smooth longitudinal trend is shown in Appendix Figure 2. In sensitivity analyses, standard multivariable linear regression models relating changes in plasma lipid concentrations over time to changes in selenium levels yielded similar results (data not shown).

Figures

Grahic Jump Location
Appendix Figure 1.
Lipid concentrations, by plasma selenium concentration at baseline.

Curves represent mean baseline lipid concentrations (solid line) and their 95% CIs (dashed line) based on restricted quadratic splines for baseline selenium concentrations, with knots at the 5th, 50th, and 95th percentiles (61.0, 87.8, and 119.0 ng/g, respectively). Results were obtained from linear mixed models with random between-participant variations in baseline lipid levels and adjusted for age, sex, center, smoking status, body mass index, and use of lipid-lowering medications (Appendix). To convert cholesterol values to mg/dL, divide by 0.02586. Scatterplots represent adjusted baseline concentrations of selenium and lipids. HDL = high-density lipoprotein.

Grahic Jump Location
Grahic Jump Location
Appendix Figure 2.
Changes in lipid concentrations after 6 months, by change in plasma selenium concentration.

Curves represent mean changes in lipid concentrations from baseline to 6 months (solid line) and their 95% CIs (dashed line) based on restricted quadratic splines for within-participant changes in selenium concentration, with knots at the 5th, 50th, and 95th percentiles (−8.0, 74.0, and 173.9 ng/g, respectively). Results were obtained from linear mixed models with random between-participant variations in both baseline lipid levels and lipid changes over time and adjusted for baseline selenium concentration, age, sex, center, smoking status, body mass index, and use of lipid-lowering medications (Appendix). Scatterplots represent adjusted changes in selenium and lipid concentrations over time for participants randomly assigned to receive placebo (gray dots) or 100 mcg (blue dots), 200 mcg (green dots), or 300 mcg (red dots) of selenium per day. To convert cholesterol values to mg/dL, divide by 0.02586. HDL = high-density lipoprotein.

Grahic Jump Location

Tables

Table Jump PlaceholderTable 1.  Descriptive Baseline Characteristics, Overall and by Group
Table Jump PlaceholderTable 2.  Effect of Selenium Supplementation on Lipid and Plasma Selenium Concentrations After 6 Months
Table Jump PlaceholderAppendix Table.  Effect of Selenium Supplementation on Changes in Lipid Concentrations After 6 Months, by Study Center
Table Jump PlaceholderTable 3.  Cross-sectional Association of Plasma Selenium Concentrations With Lipid Concentrations at Baseline
Table Jump PlaceholderTable 4.  Longitudinal Association of Changes in Plasma Selenium Concentration With Changes in Lipid Concentrations After 6 Months

References

Salonen JT, Alfthan G, Huttunen JK, Pikkarainen J, Puska P.  Association between cardiovascular death and myocardial infarction and serum selenium in a matched-pair longitudinal study. Lancet. 1982; 2:175-9.
PubMed
CrossRef
 
Nève J.  Selenium as a risk factor for cardiovascular diseases. J Cardiovasc Risk. 1996; 3:42-7.
PubMed
 
Brigelius-Flohé R, Banning A, Schnurr K.  Selenium-dependent enzymes in endothelial cell function. Antioxid Redox Signal. 2003; 5:205-15.
PubMed
 
Sattler W, Maiorino M, Stocker R.  Reduction of HDL- and LDL-associated cholesterylester and phospholipid hydroperoxides by phospholipid hydroperoxide glutathione peroxidase and Ebselen (PZ 51). Arch Biochem Biophys. 1994; 309:214-21.
PubMed
 
Salonen JT, Salonen R, Seppänen K, Kantola M, Parviainen M, Alfthan G. et al.  Relationship of serum selenium and antioxidants to plasma lipoproteins, platelet aggregability and prevalent ischaemic heart disease in Eastern Finnish men. Atherosclerosis. 1988; 70:155-60.
PubMed
 
Blankenberg S, Rupprecht HJ, Bickel C, Torzewski M, Hafner G, Tiret L, et al. AtheroGene Investigators.  Glutathione peroxidase 1 activity and cardiovascular events in patients with coronary artery disease. N Engl J Med. 2003; 349:1605-13.
PubMed
 
Vunta H, Davis F, Palempalli UD, Bhat D, Arner RJ, Thompson JT. et al.  The anti-inflammatory effects of selenium are mediated through 15-deoxy-Delta12,14-prostaglandin J2 in macrophages. J Biol Chem. 2007; 282:17964-73.
PubMed
 
Gao Y, Hannan NR, Wanyonyi S, Konstantopolous N, Pagnon J, Feng HC. et al.  Activation of the selenoprotein SEPS1 gene expression by pro-inflammatory cytokines in HepG2 cells. Cytokine. 2006; 33:246-51.
PubMed
 
Curran JE, Jowett JB, Elliott KS, Gao Y, Gluschenko K, Wang J. et al.  Genetic variation in selenoprotein S influences inflammatory response. Nat Genet. 2005; 37:1234-41.
PubMed
 
Flores-Mateo G, Navas-Acien A, Pastor-Barriuso R, Guallar E.  Selenium and coronary heart disease: a meta-analysis. Am J Clin Nutr. 2006; 84:762-73.
PubMed
 
Stranges S, Marshall JR, Trevisan M, Natarajan R, Donahue RP, Combs GF. et al.  Effects of selenium supplementation on cardiovascular disease incidence and mortality: secondary analyses in a randomized clinical trial. Am J Epidemiol. 2006; 163:694-9.
PubMed
 
Bjelakovic G, Nikolova D, Gluud LL, Simonetti RG, Gluud C.  Mortality in randomized trials of antioxidant supplements for primary and secondary prevention: systematic review and meta-analysis. JAMA. 2007; 297:842-57.
PubMed
 
Lippman SM, Klein EA, Goodman PJ, Lucia MS, Thompson IM, Ford LG. et al.  Effect of selenium and vitamin E on risk of prostate cancer and other cancers: the Selenium and Vitamin E Cancer Prevention Trial (SELECT). JAMA. 2009; 301:39-51.
PubMed
 
Stranges S, Navas-Acien A, Rayman MP, Guallar E.  Selenium status and cardiometabolic health: state of the evidence. Nutr Metab Cardiovasc Dis. 2010; 20:754-60.
PubMed
 
Jossa F, Trevisan M, Krogh V, Farinaro E, Giumetti D, Fusco G. et al.  Serum selenium and coronary heart disease risk factors in southern Italian men. Atherosclerosis. 1991; 87:129-34.
PubMed
 
Suadicani P, Hein HO, Gyntelberg F.  Serum selenium concentration and risk of ischaemic heart disease in a prospective cohort study of 3000 males. Atherosclerosis. 1992; 96:33-42.
PubMed
 
Gámez C, Ruiz-López D, Artacho R, Navarro M, Puerta A, López C.  Serum selenium in institutionalized elderly subjects and relation to other nutritional markers [Letter]. Clin Chem. 1997; 43:693-4.
PubMed
 
Coudray C, Roussel AM, Mainard F, Arnaud J, Favier A.  Lipid peroxidation level and antioxidant micronutrient status in a pre-aging population; correlation with chronic disease prevalence in a French epidemiological study (Nantes, France). J Am Coll Nutr. 1997; 16:584-91.
PubMed
 
Bleys J, Navas-Acien A, Stranges S, Menke A, Miller ER 3rd, Guallar E.  Serum selenium and serum lipids in US adults. Am J Clin Nutr. 2008; 88:416-23.
PubMed
 
Laclaustra M, Stranges S, Navas-Acien A, Ordovas JM, Guallar E.  Serum selenium and serum lipids in US adults: National Health and Nutrition Examination Survey (NHANES) 2003-2004. Atherosclerosis. 2010; 210:643-8.
PubMed
 
Stranges S, Laclaustra M, Ji C, Cappuccio FP, Navas-Acien A, Ordovas JM. et al.  Higher selenium status is associated with adverse blood lipid profile in British adults. J Nutr. 2010; 140:81-7.
PubMed
 
Yang KC, Lee LT, Lee YS, Huang HY, Chen CY, Huang KC.  Serum selenium concentration is associated with metabolic factors in the elderly: a cross-sectional study. Nutr Metab (Lond). 2010; 7:38.
PubMed
 
Bates CJ, Thane CW, Prentice A, Delves HT.  Selenium status and its correlates in a British national diet and nutrition survey: people aged 65 years and over. J Trace Elem Med Biol. 2002; 16:1-8.
PubMed
 
Salonen JT, Salonen R, Seppänen K, Kantola M, Parviainen M, Alfthan G. et al.  Relationship of serum selenium and antioxidants to plasma lipoproteins, platelet aggregability and prevalent ischaemic heart disease in Eastern Finnish men. Atherosclerosis. 1988; 70:155-60.
PubMed
 
Luoma PV, Sotaniemi EA, Korpela H, Kumpulainen J.  Serum selenium, glutathione peroxidase activity and high-density lipoprotein cholesterol—effect of selenium supplementation. Res Commun Chem Pathol Pharmacol. 1984; 46:469-72.
PubMed
 
Yu SY, Mao BL, Xiao P, Yu WP, Wang YL, Huang CZ. et al.  Intervention trial with selenium for the prevention of lung cancer among tin miners in Yunnan, China. A pilot study. Biol Trace Elem Res. 1990; 24:105-8.
PubMed
 
Rayman M, Thompson A, Warren-Perry M, Galassini R, Catterick J, Hall E. et al.  Impact of selenium on mood and quality of life: a randomized, controlled trial. Biol Psychiatry. 2006; 59:147-54.
PubMed
 
Rayman MP, Thompson AJ, Bekaert B, Catterick J, Galassini R, Hall E. et al.  Randomized controlled trial of the effect of selenium supplementation on thyroid function in the elderly in the United Kingdom. Am J Clin Nutr. 2008; 87:370-8.
PubMed
 
Bekaert B, Cooper ML, Green FR, McNulty H, Pentieva K, Scott JM. et al.  Effect of selenium status and supplementation with high-selenium yeast on plasma homocysteine and B vitamin concentrations in the UK elderly. Mol Nutr Food Res. 2008; 52:1324-33.
PubMed
 
Lide DR, ed.  CRC Handbook of Chemistry and Physics. 74th ed. Boca Raton, FL: CRC Pr; 1993-1994.
 
Wolf NM, Mueller K, Hirche F, Most E, Pallauf J, Mueller AS.  Study of molecular targets influencing homocysteine and cholesterol metabolism in growing rats by manipulation of dietary selenium and methionine concentrations. Br J Nutr. 2010; 104:520-32.
PubMed
 
Dhingra S, Bansal MP.  Hypercholesterolemia and apolipoprotein B expression: regulation by selenium status. Lipids Health Dis. 2005; 4:28.
PubMed
 
Mazur A, Nassir F, Gueux E, Moundras C, Bellanger J, Grolier P. et al.  Diets deficient in selenium and vitamin E affect plasma lipoprotein and apolipoprotein concentrations in the rat. Br J Nutr. 1996; 76:899-907.
PubMed
 
Wójcicki J, Rózewicka L, Barcew-Wiszniewska B, Samochowiec L, Juźwiak S, Kadłubowska D. et al.  Effect of selenium and vitamin E on the development of experimental atherosclerosis in rabbits. Atherosclerosis. 1991; 87:9-16.
PubMed
 
Sengupta A, Carlson BA, Hoffmann VJ, Gladyshev VN, Hatfield DL.  Loss of housekeeping selenoprotein expression in mouse liver modulates lipoprotein metabolism. Biochem Biophys Res Commun. 2008; 365:446-52.
PubMed
 
Moosmann B, Behl C.  Selenoprotein synthesis and side-effects of statins. Lancet. 2004; 363:892-4.
PubMed
 
Arnaud J, Akbaraly TN, Hininger-Favier I, Berr C, Roussel AM.  Fibrates but not statins increase plasma selenium in dyslipidemic aged patients—the EVA study. J Trace Elem Med Biol. 2009; 23:21-8.
PubMed
 
Ricote M, Li AC, Willson TM, Kelly CJ, Glass CK.  The peroxisome proliferator-activated receptor-gamma is a negative regulator of macrophage activation. Nature. 1998; 391:79-82.
PubMed
 
Touyz RM, Schiffrin EL.  Peroxisome proliferator-activated receptors in vascular biology-molecular mechanisms and clinical implications. Vascul Pharmacol. 2006; 45:19-28.
PubMed
 
Klopotek A, Hirche F, Eder K.  PPAR gamma ligand troglitazone lowers cholesterol synthesis in HepG2 and Caco-2 cells via a reduced concentration of nuclear SREBP-2. Exp Biol Med (Maywood). 2006; 231:1365-72.
PubMed
 
Xun P, Liu K, Morris JS, Daviglus ML, He K.  Longitudinal association between toenail selenium levels and measures of subclinical atherosclerosis: the CARDIA trace element study. Atherosclerosis. 2010; 210:662-7.
PubMed
 
Duffield AJ, Thomson CD, Hill KE, Williams S.  An estimation of selenium requirements for New Zealanders. Am J Clin Nutr. 1999; 70:896-903.
PubMed
 
Bleys J, Navas-Acien A, Guallar E.  Serum selenium levels and all-cause, cancer, and cardiovascular mortality among US adults. Arch Intern Med. 2008; 168:404-10.
PubMed
 
Halloran P, Roetering H, Pisani T, van den Berg B, Cobbaert C.  Reference standardization and analytical performance of a liquid homogeneous high-density lipoprotein cholesterol method compared with chemical precipitation method. Arch Pathol Lab Med. 1999; 123:317-26.
PubMed
 
Centers for Disease Control and Prevention (CDC); National Center for Health Statistics (NCHS).  National Health and Nutrition Examination Survey Data. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2003-2004. Accessed atwww.cdc.gov/nchs/nhanes/nhanes2003-2004/nhanes03_04.htmon 23 March 2011.
 
Stranges S, Marshall JR, Natarajan R, Donahue RP, Trevisan M, Combs GF. et al.  Effects of long-term selenium supplementation on the incidence of type 2 diabetes: a randomized trial. Ann Intern Med. 2007; 147:217-23.
PubMed
 
Raudenbush SW, Bryk AS.  Hierarchical Linear Models: Applications and Data Analysis Methods. 2nd ed. Thousand Oaks, CA: Sage; 2002.
 
Albert PS.  Longitudinal data analysis (repeated measures) in clinical trials. Stat Med. 1999; 18:1707-32.
PubMed
 
Greenland S.  Dose-response and trend analysis in epidemiology: alternatives to categorical analysis. Epidemiology. 1995; 6:356-65.
PubMed
 
Diggle PJ, Heagerty P, Liang KY, Zeger SL.  Analysis of Longitudinal Data. 2nd ed. Oxford, United Kingdom: Oxford Univ Pr; 2002.
 

Letters

NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Comments

Submit a Comment
Submit a Comment

Summary for Patients

The Effects of Selenium Supplements on Blood Cholesterol Levels

The full report is titled “Effect of Supplementation With High-Selenium Yeast on Plasma Lipids. A Randomized Trial.” It is in the 17 May 2011 issue of Annals of Internal Medicine (volume 154, pages 656-665). The authors are M.P. Rayman, S. Stranges, B.A. Griffin, R. Pastor-Barriuso, and E. Guallar.

Read More...

Clinical Slide Sets

Terms of Use

The In the Clinic® slide sets are owned and copyrighted by the American College of Physicians (ACP). All text, graphics, trademarks, and other intellectual property incorporated into the slide sets remain the sole and exclusive property of the ACP. The slide sets may be used only by the person who downloads or purchases them and only for the purpose of presenting them during not-for-profit educational activities. Users may incorporate the entire slide set or selected individual slides into their own teaching presentations but may not alter the content of the slides in any way or remove the ACP copyright notice. Users may make print copies for use as hand-outs for the audience the user is personally addressing but may not otherwise reproduce or distribute the slides by any means or media, including but not limited to sending them as e-mail attachments, posting them on Internet or Intranet sites, publishing them in meeting proceedings, or making them available for sale or distribution in any unauthorized form, without the express written permission of the ACP. Unauthorized use of the In the Clinic slide sets will constitute copyright infringement.

Toolkit

Want to Subscribe?

Learn more about subscription options

Advertisement
Related Articles
Journal Club
Related Point of Care
Topic Collections
PubMed Articles
Forgot your password?
Enter your username and email address. We'll send you a reminder to the email address on record.
(Required)
(Required)