Low serum albumin and the acute phase response predict low serum selenium in HIV-1 infected women
BMC Infectious Diseases
Low serum albumin and the acute phase response predict low serum selenium in HIV-1 infected women
Paul K Drain 2
Jared M Baeten 1
Julie Overbaugh 3
Mark H Wener 0 1 6
Daniel D Bankson 0 4 6
Ludo Lavreys 5 7
Kishorchandra Mandaliya 8
Jeckoniah O Ndinya-Achola 7
R Scott McClelland 1 5 7
0 Department of Laboratory Medicine, University of Washington, 1959 NE Pacific, A-300 Health Sciences , Box 356340, Seattle, WA 98105 , USA
1 Department of Medicine, University of Washington, 1959 NE Pacific, A-300 Health Sciences , Box 356340, Seattle, WA 98105 , USA
2 School of Medicine, University of Washington, 1959 NE Pacific, A-300 Health Sciences , Box 356340, Seattle, WA 98105 , USA
3 Divisions of Human Biology and Public Health Sciences, Fred Hutchinson Cancer Research Center , Seattle , USA
4 Veterans Affairs Puget Sound Health Care System , Seattle , USA
5 Department of Epidemiology, University of Washington, 1959 NE Pacific, A-300 Health Sciences , Box 356340, Seattle, WA 98105 , USA
6 Clinical Nutrition Research Unit Laboratory Core, University of Washington, 1959 NE Pacific, A-300 Health Sciences , Box 356340, Seattle, WA 98105 , USA
7 Department of Medical Microbiology, University of Nairobi , Nairobi , Kenya
8 Department of Pathology, Coast Provincial General Hospital , Mombasa , Kenya
Background: Low serum selenium has been associated with lower CD4 counts and greater mortality among HIV-1-seropositive individuals, but most studies have not controlled for serum albumin and the presence of an acute phase response. Methods: A cross-sectional study was conducted to evaluate relationships between serum selenium concentrations and CD4 count, plasma viral load, serum albumin, and acute phase response markers among 400 HIV-1-seropositive women. Results: In univariate analyses, lower CD4 count, higher plasma viral load, lower albumin, and the presence of an acute phase response were each significantly associated with lower serum selenium concentrations. In multivariate analyses including all four of these covariates, only albumin remained significantly associated with serum selenium. For each 0.1 g/dl increase in serum albumin, serum selenium increased by 0.8 g/l (p < 0.001). Women with an acute phase response also had lower serum selenium (by 5.6 g/l, p = 0.06). Conclusion: Serum selenium was independently associated with serum albumin, but not with CD4 count or plasma viral load, in HIV-1-seropositive women. Our findings suggest that associations between lower serum selenium, lower CD4 count, and higher plasma viral load may be related to the frequent occurrence of low serum albumin and the acute phase response among individuals with more advanced HIV-1 infection.
* Corresponding author
Nutritional deficiencies have long been recognized as an
important problem among HIV-1-seropositive
individuals, particularly in resource-limited settings .
Micronutrient deficiencies have been associated with more rapid
HIV-disease progression and higher HIV-1 related
mortality [2,3]. In some studies, micronutrient supplementation
has delayed time to AIDS and improved survival,
suggesting that supplementation could offer a simple and
relatively inexpensive strategy to slow HIV-1 progression
Selenium is an antioxidant micronutrient that is an
essential element of selenoproteins, including selenoprotein P
and glutathione peroxidase. Among HIV-1-seropositive
individuals, lower serum selenium concentrations have
been associated with lower CD4 counts, more advanced
HIV-1 disease, and greater HIV-1 related mortality [6-10].
However, most studies have not controlled for low serum
albumin, which binds non-specifically to selenium in
serum, or for the presence of an acute phase response,
which alters hepatic production of albumin and other
serum proteins [11,12]. We sought to determine whether
serum selenium was independently associated with CD4
count or plasma viral load after adjusting for serum
albumin and the presence of an acute phase response.
A cross-sectional study was conducted using baseline data
from 400 HIV-1-seropositive women enrolled in a
randomized trial of micronutrient supplementation . Data
were collected between September 1998 and June 2000 at
Coast Provincial General Hospital in Mombasa, Kenya.
Women between 18 and 45 years old were enrolled if they
were not currently or recently (last 3 months) pregnant,
taking vitamin supplements, or using oral contraceptives.
The enrollment criteria were based on the parent trial of
micronutrient supplementation . All participants
were antiretroviral nave. The protocol was approved by
the institutional review boards of the University of
Nairobi and the University of Washington, and all women
provided written informed consent.
Detailed procedures and sample collection techniques
have been previously described . In brief, women were
interviewed regarding demographic, sexual, and medical
characteristics using a standardized questionnaire. A
physical examination was performed. Blood was collected for
lymphocyte subset analysis, quantitation of plasma
1 RNA, and nutritional assays.
Serum samples were protected from light, separated
within 4 hours of collection, and stored at -70C.
Serological testing for HIV-1 was performed using an ELISA
(Detect HIV 1/2, BioChem Immunosystems, Montreal,
Canada), and confirmed with a second ELISA
(Recombigen, Cambridge Biotech, Worcester, USA). Absolute
CD4 counts were determined using a semiautomated
system (Zymmune CD4/CD8 Cell Monitoring Kit, Bartels
Inc., Issaquah, USA), which had a lower quantitation
limit of 25 cells/l. The quantity of HIV-1 RNA in plasma
was determined using the Gen-Probe HIV-1 viral load
assay (Gen-Probe Incorporated, San Diego, USA). The
lower limit of quantification for the assay was 3 copies/
reaction, which was equivalent to 12 copies/ml for the
plasma volumes tested . Serum selenium was
quantified using graphite furnace atomic absorption
spectrophotometry . Serum albumin, C-reactive protein
(CRP), and 1-acid glycoprotein (AGP) were determined
by nephelometry (Dade Behring, Marburg, Germany).
Data were analyzed using SPSS 12.0 (SPSS Inc., Chicago,
USA). Low serum selenium was defined as a serum level
85 g/l , a threshold that has been associated with
adverse outcomes in HIV-1 infection [7,8]. An acute phase
response was considered to be present if a participant had
CRP 1 mg/dl  or AGP 100 mg/dl . Univariate
comparisons were performed using chi-square tests for
dichotomous outcomes and t-tests for continuous
outcomes. Multivariate comparisons were conducted using
logistic and linear regression. Plasma HIV-1 RNA levels
were log10 transformed to approximate a normal
Baseline characteristics of this study population have been
described . In brief, participants had a mean age of 29
years [standard deviation (SD) 6] with 7 years (SD 4)
of education. Participants were generally of low
socioeconomic status, as evidenced by only 35 (9%) having a toilet
in the home. Two hundred twenty (55%) participants
were married. The women had a mean of 3 children (SD
2), and 73 (18%) were using injectable progesterone
contraception (depot medroxyprogesterone acetate). The
mean serum selenium concentration was 100 g/l (SD
Comparison of women with low vs. normal serum selenium
Women with low serum selenium had more advanced
immunosuppression, higher plasma viral loads, lower
albumin, more frequent symptoms and signs of HIV-1
infection, and were more likely to have an acute phase
response compared to women with normal serum
selenium (Table 1). In a multivariate model including CD4
count, plasma viral load, albumin, and the acute phase
response, only serum albumin concentration and the
Mean ( SD) or Number (%)
Multivariate Logistic Regression1
Adjusted Odds Ratio (95% CI)
SD standard deviation; CI confidence interval
1 CD4 count, plasma HIV-1 RNA, serum albumin were modeled as continuous variables, and the acute phase response was modeled as a
2 Calculated by using t-tests for continuous variables and 2 tests for dichotomous variables.
3 Odds ratio is per 100 CD4 cells/l increase.
4 The presence of C-reactive protein 1 mg/dl and/or 1-acid glycoprotein 100 mg/dl.
5 Defined as fever for 1 month, diarrhea for 1 month, cough for 1 month, unintended weight loss of 5 kg during previous year, or itching
skin rash during previous year.
6 Defined as the presence of oral thrush, oral hairy leukoplakia, oral ulcer, maculopapular rash, or Kaposi's sarcoma.
presence of an acute phase response remained
significantly associated with low serum selenium. When
included in the multivariate model, signs and symptoms
of HIV-1 and body mass index were not independently
associated with serum selenium and their inclusion did
not significantly affect the association between selenium
and albumin or the acute phase response, so these
covariates were not included in the final multivariate model.
Correlates of serum selenium
In univariate analyses, higher CD4 count and higher
serum albumin concentrations were associated with
higher serum selenium concentrations, while higher
plasma viral load, the presence of an acute phase
response, and symptoms or signs of HIV-1 disease were
associated with lower selenium concentrations (Table 2).
In a multivariate model including CD4 count, plasma
viral load, albumin, and the acute phase response, only
albumin was significantly associated with serum
selenium. Each 0.1 g/dl increase in serum albumin was
associated with an 0.8 g/l [95% confidence interval (CI) 0.4
1.2] increase in serum selenium. Women with an acute
phase response had lower serum selenium concentrations
than women without an acute phase response, although
this association did not reach statistical significance. Signs
or symptoms of HIV-1 and body mass index were not
associated with serum selenium and did not substantially
affect the associations between selenium and albumin or
the acute phase response results, so these variables were
not included in the final multivariate model. In separate
multivariate models evaluating CRP and AGP as
continuous covariates, neither of these inflammatory markers was
Univariate Linear Regression
Multivariate Linear Regression
CI confidence interval
1The presence of C-reactive protein 1 mg/dl and/or 1-acid glycoprotein 100 mg/dl.
2 Defined as fever for 1 month, diarrhea for 1 month, cough for 1 month, unintended weight loss of 5 kg during previous year, or itching
skin rash during previous year.
3 Defined as the presence of oral thrush, oral hairy leukoplakia, oral ulcer, maculopapular rash, or Kaposi's sarcoma.
In this cross-sectional study of HIV-1-seropositive
women, low serum selenium was independently
associated with serum albumin and with the acute phase
response, but not with CD4 count or plasma viral load.
Further prospective studies may help determine whether
associations between low serum selenium and low CD4
count [6,9] and more advanced HIV-1 disease  could
be related to the frequent occurrence of
hypoalbuminemia and the acute phase response in people with
advanced HIV-1 infection.
Several ingested forms of selenium, including
selenomethionine, bind non-specifically to albumin for transport
to the liver [11,18-21]. The liver converts these
compounds into selenocysteine, which is used to form various
selenoproteins. In total, approximately 55% of selenium
in human serum exists in selenoprotein P, another 17
32% exists bound to albumin, mostly in the form of
selenomethionine, and only 10% of serum selenium is
not protein bound [11,18,21]. Since low serum albumin
has been independently associated with faster HIV-1
disease progression and higher mortality, low serum
selenium may simply reflect a decline in serum albumin
among people with more active or advanced HIV-1
The presence of an acute phase response is typically
associated with a decrease of serum albumin and other plasma
proteins . Among HIV-1-seropositive individuals, the
acute phase response has also been associated with low
serum selenium and with HIV-1 disease progression and
mortality [6,24]. One study found that CRP predicts
mortality in HIV-1-infected women independent of serum
albumin . Our results suggest that the observed
univariate associations between serum selenium and the
acute phase response may have been due, at least in part,
to decreased hepatic production of albumin and other
plasma proteins in HIV-1-seropositive individuals with an
acute phase response [18,22]. There may also be a
redistribution of selenium from serum and liver to muscle tissue
during an acute phase response .
Our study builds on previous analyses by examining the
relationship between serum selenium concentrations,
CD4 count, and plasma HIV-1 viral load in a large cohort
of untreated HIV-1-seropositive adults. The size of this
study enhanced our ability to conduct detailed
multivariate analyses, which demonstrate the lack of a significant
independent association between selenium and CD4 cell
count or plasma viral load.
We have previously published the results of the
micronutrient supplementation trial in which these women
received six weeks of either a supplement containing B
vitamins, vitamin C, vitamin E, and selenium or an
identical placebo . Following supplementation, women
who received the supplement had slightly higher CD4
counts compared to those who received placebo, an effect
that was also observed in a trial of an otherwise identical
supplement that did not contain selenium . It is not
possible to disentangle the independent effects of
selenium from the known effects of those other
micronutrients that were provided in the same supplement. Thus, we
were unable to use those longitudinal data to evaluate the
associations between selenium supplementation and
albumin, CD4 count, and plasma viral load.
The findings presented here should be interpreted in the
context of the limitations of this study. Although
crosssectional studies are useful to define associations, it is not
possible to infer with certainty that low albumin or the
acute phase response were the cause of low measured
serum selenium, although this relationship seems
plausible because a large proportion of serum selenium is
protein bound [18,22]. Regardless of the mechanism, the
confounding bias demonstrated by our analyses was
strong enough to nullify highly significant univariate
associations between serum selenium and CD4 count and
plasma viral load. However, these data cannot rule out the
possibility that low serum selenium or a low antioxidant
status was the cause of low serum albumin. Furthermore,
because hypoalbuminemia may influence the
relationship between serum selenium and total body selenium
status, the measured serum selenium may not accurately
reflect total body selenium in advanced HIV-1 infection.
Data on dietary selenium intake were not collected in this
population. Finally, because this study included only
women, these results may not be generalizable to
The finding that serum selenium is not independently
associated with CD4 count or plasma viral load may help
to explain the results of small randomized and
non-randomized trials of selenium supplementation among
HIV-1seropositive individuals. While one study found an
increase in CD4/CD8 ratio after 12 weeks , none have
demonstrated significant effects on the absolute CD4 cell
count or plasma viral load [10,26,27]. However, a
beneficial effect of selenium supplementation that is
independent of CD4 count and plasma viral load is possible. In one
randomized trial, selenium supplementation decreased
hospital admissions due to infections among HIV-1
infected adults . The trial did not report changes in
biological markers of HIV-1 disease progression or the
effect on HIV-1-related mortality.
The results of this investigation demonstrate that serum
selenium was not independently associated with CD4
count or plasma viral load among HIV-1-seropositive
women. These findings indicate that studies assessing the
impact of selenium on HIV-1 surrogate markers, such as
CD4 cell count and plasma viral load, need to control for
serum albumin levels and the presence of an acute phase
AGP 1-acid glycoprotein
CI Confidence Interval
CRP C-reactive protein
SD Standard Deviation
PKD, JMB, KM, JON, and RSM designed the study. LL,
JMB, KM and RSM collected data and provided study
oversight. PKD and RSM analyzed data. PKD, JMB, JO, MHW,
DDB, and RSM interpreted the results. PKD and RSM
primarily wrote the manuscript. JMB, JO, MHW, DDH
provided valuable insight for revising the manuscript. All
authors read and approved the final manuscript.
This research was supported by National Institutes of Health grants
AI43844 and AI39996, and University of Washington Clinical Nutrition
Research Unit grant DK35816. JM Baeten was supported by the Fogarty
International Center (FIC) grant D43-TW00007. RS McClelland was
supported by FIC D43-TW00007 and by K23-AI52480. PK Drain was
supported by the David E. Rogers Fellowship of The New York Academy of
The authors wish to acknowledge the excellent work and valuable
contributions made to this study by our clinic (Mary Wamugunda, Virginia Njuki,
and Florence Murigi) and laboratory staff (Bhavna Chohan, Khamis
Mwinyikai, Amina Abdalla, Gladwell Maina, Sandra Emery, and Dana Panteleeff).
Del Landicho provided assistance with nutritional testing performed at the
University of Washington's Clinical Nutrition Research Unit. We thank
Coast Provincial General Hospital for allowing us to use their clinical
facilities. Finally, we would like to express our gratitude to the women who
participated in this study, without whose time and effort this research would
not have been possible.
Beach RS , Mantero-Atienza E , Shor-Posner G , Javier JJ , Szapocznik J , Morgan R , Sauberlich HE , Cornwell PE , Eisdorfer C , Baum MK : Specific nutrient abnormalities in asymptomatic HIV-1 infection . AIDS 1992 , 6 : 701 - 708 .
Semba RD , Graham NM , Caiaffa WT , Margolick JB , Clement L , Vlahov D : Increased mortality associated with vitamin A deficiency during human immunodeficiency virus type 1 infection . Arch Intern Med 1993 , 153 : 2149 - 2154 .
3. Tang AM , Graham NM , Semba RD , Saah AJ : Association between serum vitamin A and E levels and HIV-1 disease progression . AIDS 1997 , 11 : 613 - 620 .
4. Jiamton S , Pepin J , Suttent R , Filteau S , Mahakkanukrauh B , Hanshaoworakul W , Chaisilwattana P , Suthipinittharm P , Shetty P , Jaffar S : A randomized trial of the impact of multiple micronutrient supplementation on mortality among HIV-infected individuals living in Bangkok . AIDS 2003 , 17 : 2461 - 2469 .
5. Fawzi WW , Msamanga GI , Spiegelman D , Wei R , Kapiga S , Villamor E , Mwakagile D , Mugusi F , Hertzmark E , Essex M , Hunter DJ : A randomized trial of multivitamin supplements and HIV disease progression and mortality . N Engl J Med 2004 , 351 : 23 - 32 .
6. Look MP , Rockstroch JK , Rao GS , Kreuzer KA , Spengler U , Sauerbruch T : Serum selenium versus lymphocyte subsets and markers of disease progression and inflammatory response in human immunodeficiency virus-infection . Biol Trace Elem Res 1997 , 56 : 31 - 41 .
7. Campa A , Shor-Posner G , Indacochea F , Zhang G , Lai H , Asthana D , Scott GB , Baum MK : Mortality risk in selenium-deficient HIVpositive children . J Acquir Immune Defic Syndr 1999 , 20 : 508 - 513 .
8. Baum MK , Shor-Posner G , Lai S , Zhang G , Lai H , Fletcher MA , Sauberlich H , Page JB : High risk of HIV-related mortality is associated with selenium deficiency . J Acquir Immune Defic Syndr Hum Retrovirol 1997 , 15 : 370 - 374 .
9. Kupka R , Msamanga GI , Spiegelman D , Morris S , Mugusi F , Hunter DJ , Fawzi WW : Selenium status is associated with accelerated HIV disease progression among HIV-1-infected pregnant women in Tanzania . J Nutr 2004 , 134 : 2556 - 2560 .
10. Cirelli A , Ciardi M , De Simone C , Sorice F , Giordano R , Ciaralli L , Constantini S : Serum selenium concentration and disease progress in patients with HIV infection . Clin Biochem 1991 , 24 : 211 - 214 .
11. Deagan JT , Butler JA , Zachara A , Whanger PD : Determination of the distribution of selenium between glutathione peroxidase, selenoprotein P, and albumin in plasma . Analytical Biochem 1993 , 208 : 176 - 181 .
12. Gabay C , Kushner I : Acute-phase proteins and other systemic responses to inflammation . N Eng J Med 1999 , 340 : 448 - 454 .
13. McClelland RS , Baeten JM , Overbaugh J , Richardson BA , Mandaliya K , Emery S , Lavreys L , Ndinya-Achola JO , Bankson DD , Bwayo JJ , Kreiss JK : Multivitamin supplementation increases genital tract shedding of HIV-1 in women: Results of a randomized trial . J Acquir Immune Defic Syndr 2004 , 37 : 1657 - 1663 .
14. Panteleeff DD , Emery S , Richardson BA , Rousseau C , Benki S , Bodrug S , Kreiss JK , Overbaugh J : Validation of performance of the GenProbe human immunodeficiency virus type 1 viral load assay with genital swabs and breast milk samples . J Clin Microbiol 2002 , 40 : 3929 - 3937 .
15. Ericson SP , McHalsky ML , Rabinow BE , Kronholm KG , Arceo CS , Weltzer JA , Ayd SW : Sampling and analysis techniques for monitoring serum for trace elements . Clin Chem 1986 , 32 : 1350 - 1356 .
16. Hardy G , Hardy I : Selenium: The Se-XY nutraceutical . Nutrition 2004 , 20 : 590 - 593 .
17. Thurnham DI , McCabe GP , Northrop-Clewes , Nestel P : Effects of subclinical infection on plasma retinol concentrations and assessment of prevalence of vitamin A deficiency: meta-analysis . Lancet 2003 , 362 : 2052 - 2058 .
18. Harrison I , Littlejohn D , Fell GS : Distribution of selenium in human blood plasma and serum . Analyst 1996 , 121 : 189 - 194 .
19. Shiobara Y , Suzuki KT : Binding of selenium (administered as selenite) to albumin after efflux from red blood cells . J Chromatogr B Biomed Sci Appl 1998 , 710 : 49 - 56 .
20. Burk RF , Hill KE , Motley AK : Plasma selenium in specific and non-specific forms . Biofactors 2001 , 14 : 107 - 114 .
21. Suzuki KT , Itoh M : Metabolism of selenite labelled with enriched stable isotope in the bloodstream . J Chromatogr B Biomed Sci Appl 1997 , 692 : 15 - 22 .
22. Maehira F , Luyo GA , Miyagi I , Oshiro M , Yamane N , Kuba M , Nakazato Y : Alterations of serum selenium concentrations in the acute phase of pathological conditions . Clin Chim Acta 2002 , 316 : 137 - 146 .
23. Feldman JG , Gange SJ , Bacchetti P , Cohen M , Young M , Squires KE , Williams C , Goldwasser P , Anastos K : Serum albumin is a powerful predictor of survival in HIV-1-infected women . J Acquir Immune Defic Syndr 2003 , 33 : 66 - 73 .
24. Baeten JM , McClelland RS , Richardson BA , Bankson DD , Lavreys L , Wener MH , Overbaugh J , Mandaliya K , Ndinya-Achola JO , Bwayo JJ , Kreiss JK : Vitamin A deficiency and the acute phase response among HIV-1-infected and -uninfected women in Kenya . J Acquir Immune Defic Syndr 2002 , 31 : 243 - 249 .
25. Feldman JG , Goldwasser P , Holman S , DeHovitz J , Minkoff H : Creactive protein is an independent predictor of mortality in women with HIV-1 infection . J Acquir Immune Defic Syndr 2003 , 32 : 210 - 214 .
26. Look MP , Rockstroh JK , Rao GS , Barton S , Lemoch H , Kaiser R , Kupfer B , Sudhop T , Spengler U , Sauerbruch T : Sodium selenite and N-acetylcysteine in antiretrovial-naive HIV-1-infected patients: a randomized, controlled pilot study . Euro J Clin Investigation 1998 , 28 : 389 - 397 .
27. Delmas-Beauvieux MC , Peuchant E , Couchouron A , Constans J , Sergeant C , Simonoff M , Pellegrin JL , Leng B , Conri C , Clerc M : The enzymatic antioxidant system in blood and glutathione status in human immunodeficiency virus (HIV)-infected patients: effects of supplementation with selenium or -carotene . Am J Clin Nutr 1996 , 64 : 101 - 107 .
28. Burbano X , Miguez-Burbano MJ , McCollister K , Zhang G , Rodriguez A , Ruiz P , Lecusay R , Shor-Posner G : Impact of a selenium chemoprevention clinical trial on hospital admissions of HIVinfected participants . HIV Clin Trials 2002 , 3 : 483 - 491 .