Statement of Significance
To the best of our knowledge, this is the first study to review systematically, with a comprehensive analysis, results of clinical trials on Se supplementation in patients with chronic kidney disease undergoing haemodialysis.
Introduction
Selenium (Se) is an essential trace element that exerts important functions for human health through several selenoproteins(Reference Avery and Hoffmann1,Reference Hariharan and Dharmaraj2) . Selenium is mainly consumed via diet, with Brazil nuts, cereals, seeds, green vegetables, fish, seafood, beef and poultry being the main food sources(Reference Prabhu and Lei3).
Selenium is co-translationally incorporated as selenocysteine into the polypeptide chains of the selenoproteins by a UGA codon(Reference Papp, Lu and Holmgren4,Reference Bulteau and Chavatte5) . A total of twenty-five selenoproteins have been identified in humans so far, including several families such as glutathione peroxidases (GPx), the main antioxidant family enzyme; thioredoxin reductases (TrxR), which maintain the redox status of intracellular proteins and act as apoptosis inhibitors; and iodothyronine deiodinases (DIO), which regulate the activation or inactivation of thyroid hormones by reductive deiodination. Further, selenoprotein P (SelenoP) is the main selenium transporter owing to its capacity to carry up to ten selenocysteine residues, besides playing a role in redox signalling(Reference Papp, Lu and Holmgren4,Reference Burk and Hill6–Reference Saito10) . Se activities in the body depend on its adequate intake; thus, its deficiency, even moderate, can compromise the synthesis of selenoproteins and contribute to increased oxidative stress, poorer immune system response, increased risk of infections, cardiovascular and neurodegenerative diseases, and thyroid dysfunction(Reference Hariharan and Dharmaraj2,Reference Papp, Lu and Holmgren4,Reference Ventura, Melo and Carrilho11–Reference Rayman13) .
Studies have shown that patients with chronic kidney disease (CKD) have a greater propensity to have vitamin and mineral deficiency aggravated by reduced food intake, chronic inflammation and oxidative stress present in this condition(Reference Karkar14–Reference Mihai, Codrici and Popescu17). Among the minerals, the literature reports that patients with CKD commonly present reduced Se levels compared with healthy adults(Reference Gómez de Oña, Martínez-Morillo and Gago González18–Reference Almeida, Gajewska and Duro22). Se deficiency is particularly highlighted in CKD as it can contribute to a reduced antioxidant response, worse nutritional and clinical status of patients(Reference Karkar14–Reference Mihai, Codrici and Popescu17) and metabolic complications such as oxidative stress and inflammation, alteration in thyroid hormones, cardiovascular disease, infectious diseases and high mortality risk(Reference Fujishima, Ohsawa and Itai19,Reference Wang, Lin and Hsu23–Reference Anadón Ruiz, Martín Jiménez and Bermejo-Barrera27) .
Furthermore, patients undergoing haemodialysis (HD) exhibit higher levels of inflammation and oxidative stress, since the biocompatibility of dialysis membranes can lead to leucocyte activation and the production of inflammatory cytokines and pro-oxidants(Reference Descamps-Latscha, Drüeke and Witko-Sarsat28–Reference Ogunleye, Akinbodewa and Adejumo31). The inflammatory condition established in these patients could lead to a decrease in plasma Se resulting from the redistribution of Se circulating(Reference Berger, Shenkin and Schweinlin32). Pro-inflammatory cytokine production can increase capillaries’ permeability allowing transport proteins to flow through it, such as albumin, and suppress the production of transport proteins in the liver, such as selenoprotein P. Moreover, the kidneys are identified as the main source of plasma GPx synthesis(Reference Avissar, Ornt and Yagil33), so changes in plasma levels of micronutrients may be due to redistribution and increased utilisation of tissues caused by oxidative stress(Reference Duncan, Talwar and McMillan34,Reference Stefanowicz, Talwar and O’Reilly35) .
Excessive Se intake can also lead to adverse health effects, such as selenosis, that could cause tachycardia, nausea, diarrhoea, alopecia, skin lesions, fatigue and decreased cognitive function(Reference Turck and Bohn36); even high exposure at low doses may be associated with an increased risk of type 2 diabetes(Reference Vinceti, Filippini and Rothman37). Therefore, to avoid these risks, it is important to consider the tolerable upper daily intake (UL). According to the US Food and Nutrition Board of the Institute of Medicine, the recommended UL for adults is 400 μg/d(38,Reference Cheng39) . Further, the European Food Safety Authority suggests lowering the recommended UL for Se(Reference Turck and Bohn36).
Concentration of Se in whole blood, plasma or serum is the main indicator of its status; therefore, in cases of inflammation like in patients with CKD, the evaluation of erythrocyte Se associated with C-reactive protein measurement should also be considered(Reference Berger, Shenkin and Schweinlin32). According to The National Kidney Foundation’s Kidney Disease Outcomes Quality Initiative (KDOQI)(Reference Ikizler, Burrowes and Byham-Gray40), there are no specific dietary Se recommendations for patients with CKD, instead following recommendations for the general population (for example, the Recommended Dietary Allowance (RDA)).
In that regard, studies with Se supplementation in these patients are being developed to verify and analyse its effects on health outcomes. Therefore, this work proposed to systematically review the effects of Se supplementation in plasma Se levels, antioxidant and inflammatory parameters, immune system response and thyroid hormones in patients with CKD undergoing HD. The findings of this review will help to identify knowledge gaps in the effects of Se on metabolic outcomes in CKD undergoing HD to provide fundamental knowledge for nutritional management for these patients.
Methods
This systematic review followed the recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) (Supplementary Table 1)(Reference Page, McKenzie and Bossuyt41,Reference Page, Moher and Bossuyt42) , and the protocol of this review was registered in the International Prospective Register of Systematic Reviews (PROSPERO) by the registration number CRD42021231444. The PICOS strategy was used to define the research question (Table 1).
Table 1. PICOS strategy
Search strategy
The electronic databases Medline (PubMed), Lilacs, Embase, Scopus and Cochrane Library were searched from inception through July 2021, with no restrictions on publication year. The search was updated in July 2024. Search strategies were used to identify relevant scientific material (Supplementary Figure 1), combining the following terms: (Selenium) AND (‘Renal Insufficiencies Chronic’ OR ‘Kidney Failure Chronic’ OR ‘Chronic Kidney Disease’ OR Chronic Renal Disease OR ‘renal dialysis’ OR Hemodialysis OR ‘Dialyses Renal’). The filter ‘document type’ (article and review) was used in the Scopus database, and the filter ‘lilacs’ was used in the Lilacs database.
Fig. 1. Flow diagram of the selection process and studies included in the systematic review.
Inclusion and exclusion criteria
Clinical trial studies included in this review were carried out in adults over 18 years old with CKD on HD, with at least one of the following outcomes: plasma Se levels (primary outcome), antioxidant, inflammatory or immune response parameters, or thyroid hormones concentration (secondary outcomes). In addition, selenium treatment must have been performed through oral administration of supplements or rich/fortified foods. Studies in children and adolescents involving supplementation combined with other nutrients/compounds, reviews, protocols, abstracts, editorials and conference posters were excluded. Articles were eligible for inclusion if published in English, Portuguese or Spanish.
Screening
After the exclusion of duplicates, the identified articles were screened by title and abstract independently by two authors (P.C.T. and V.O.L.) using an electronic spreadsheet. Then, the selected articles underwent full-text screening by the same authors independently. In both steps, conflicts were resolved by a third author (M.B.S.P.). The articles that met the eligibility criteria were included.
Data extraction
Data were independently extracted by two authors (P.C.T. and V.O.L.), and conflicts were resolved by a third author (M.B.S.P.). The following data were extracted and recorded using an electronic spreadsheet: first author, year of publication, the country in which the study was conducted, study design, the participants’ characteristics (number of participants, age, sex, time of HD and body mass index), Se treatment (dose, type and duration), plasma Se levels at baseline, adverse effects attributable to the treatment, and the outcomes found in each study divided into primary (Se concentration in plasma) and secondary outcomes (antioxidant, inflammatory or immune response parameters, or thyroid hormones concentration). Data not available but necessary in eligible studies were requested by the authors. After the extraction, a data synthesis was performed, and the results were grouped according to the outcomes.
Risk-of-bias assessment
The risk of bias was assessed using the Cochrane risk-of-bias tool version 2 (RoB 2)(Reference Sterne, Savović and Page43), which involves assigning a low risk, some concerns of risk or a high risk of bias to the studies according to the five assessed domains: bias arising from the randomisation process; bias due to deviations from intended interventions; bias due to missing outcome data; bias in the measurement of the outcome; and bias in the selection of the reported result. Risk of bias assessment was performed independently by two reviewers (P.C.T. and V.O.L.), and conflicts were resolved by a third author (M.B.S.P.).
Results
A total of 889 records were identified in the initial search. After removing the duplicates (385), 504 records were screened by title and abstract. A total of 490 papers were excluded according to the inclusion/exclusion criteria, and the remaining fourteen articles were assessed for eligibility. In the search update, 245 records were screened by title and abstract, and five articles were assessed for eligibility. A total of nineteen studies underwent full-text screening. A total of thirteen studies met the eligibility criteria and were included in this review(Reference Bonomini, Forster and De Risio44–Reference Assarzadeh, Vahdat and Seirafian56) (Figure 1).
Study characteristics and interventions
This review included eleven randomised clinical trials(Reference Napolitano, Bonomini and Bomba45,Reference Temple, Smith and Cockram46,Reference Zachara, Gromadzinska and Zbrog48–Reference Assarzadeh, Vahdat and Seirafian56) and two non-randomised placebo control clinical trials(Reference Bonomini, Forster and De Risio44,Reference Zachara, Adamowicz and Trafikowska47) . The studies included involved a total of 722 HD patients from both sexes aged between 39.5 and 61.5 years old. Most studies (n = 8) performed the intervention for 3 months, four studies for 6 months(Reference Bonomini, Forster and De Risio44,Reference Napolitano, Bonomini and Bomba45,Reference Salimian, Soleimani and Bahmani53,Reference Jamal, Seifati and Dehghani Ashkezari54) and one for 14 d(Reference Temple, Smith and Cockram46). The daily doses of Se varied between 134 and 200 µg/d, and four studies supplemented only three times a week with doses that ranged between 200 and 500 µg/d(Reference Bonomini, Forster and De Risio44,Reference Napolitano, Bonomini and Bomba45,Reference Zachara, Adamowicz and Trafikowska47,Reference Atapour, Vahdat and Hosseini55) . One study did not report the amount of Se supplemented(Reference Omrani, Rahimi and Nikseresht52) (Table 2).
Table 2. Information and results from studies included that evaluated the effects of Se supplementation in patients with CKD on HD
↔, no difference; BMI, body mass index; CG, control group; CKD, chronic kidney disease; F, female; FT3, free triiodothyronine; CRP, C-reactive protein; GPx, glutathione peroxidase; GSH, reduced glutathione; HD, haemodialysis; hs-CRP, high sensitivity C-reactive protein; IL-1, interleukin-1; IL-6, interleukin 6; IL-8, interleukin-8; M, male; MDA, malondialdehyde; n, sample number; PPAR-γ, peroxisome proliferator-activated receptor-gamma; rT3, reverse T3; Se, selenium; SG, supplemented group; T3, triiodothyronine; T3RU, T3 resin uptake; T4, thyroxine; TGF-β, transforming growth factor-β; TNF-α, tumour necrosis factor-α; TSH, thyroid-stimulating hormone.
None of the studies used food as a source of Se; all of them performed the intervention with a supplement. Two studies carried out the supplementation using sodium selenite(Reference Bonomini, Forster and De Risio44,Reference Napolitano, Bonomini and Bomba45) , four studies supplemented in the form of Se-enriched yeast(Reference Zachara, Adamowicz and Trafikowska47–Reference Zachara, Gromadzinska and Palus49,Reference Jamal, Seifati and Dehghani Ashkezari54) and one study supplemented with an oral nutritional formula with sodium selenite or selenate(Reference Temple, Smith and Cockram46). Six studies provided limited information on the type of Se compound used for supplementation, reporting only the use of oral capsules, pills or tablets(Reference Salehi, Sohrabi and Ekramzadeh50–Reference Salimian, Soleimani and Bahmani53,Reference Atapour, Vahdat and Hosseini55,Reference Assarzadeh, Vahdat and Seirafian56) (Table 2).
Four studies reported no side effects or adverse effects attributable to Se toxicity after the intervention(Reference Bonomini, Forster and De Risio44,Reference Napolitano, Bonomini and Bomba45,Reference Salimian, Soleimani and Bahmani53,Reference Jamal, Seifati and Dehghani Ashkezari54) , and only one reported gastrointestinal disturbance (dyspepsia and abdominal pain, nausea and vomiting)(Reference Salehi, Sohrabi and Ekramzadeh50).
Risk-of-bias assessment
All the studies were classified as high risk of bias because they were identified as presenting a high risk of bias in at least one of the domains. The judgements of each risk-of-bias domain for each included study are in Supplementary Figure 2, and the percentage of each risk-of-bias domain of all included studies can be verified in Figure 2.
Fig. 2. Risk-of-bias graph about the percentages of risk of bias in evaluated clinical studies.
Five studies had some concerns for domain one (randomisation process) because they did not inform how it was performed(Reference Temple, Smith and Cockram46,Reference Zachara, Gromadzinska and Zbrog48,Reference Zachara, Gromadzinska and Palus49) or because they are non-randomised studies(Reference Bonomini, Forster and De Risio44,Reference Zachara, Adamowicz and Trafikowska47) . Only one study presented a high risk in this domain for giving a significant difference in the size of the groups, in addition to not providing details on how the randomisation was performed(Reference Napolitano, Bonomini and Bomba45).
All the studies showed a high risk of bias in deviations from the intended interventions because they did not report the type of analysis used to estimate the assignment effect to intervention by not reporting the missing data or follow-up losses. The studies by Salehi et al. (2013)(Reference Salehi, Sohrabi and Ekramzadeh50), Salimian et al. (2022)(Reference Salimian, Soleimani and Bahmani53), Jamal et al. (2022)(Reference Jamal, Seifati and Dehghani Ashkezari54) and Atapour et al. (2022)(Reference Atapour, Vahdat and Hosseini55) did not present a flow diagram of the clinical trial. Furthermore, six studies showed losses or exclusions of participants more significant than 5%, which could impact the final result(Reference Salehi, Sohrabi and Ekramzadeh50,Reference Omrani, Rahimi and Nikseresht52–Reference Assarzadeh, Vahdat and Seirafian56) .
Regarding the ‘missing outcome data’ domain, seven studies presented a low risk of bias. Despite not reporting the extent of missing data in most of them, the losses seemed unrelated to the intervention(Reference Bonomini, Forster and De Risio44,Reference Zachara, Adamowicz and Trafikowska47,Reference Salehi, Sohrabi and Ekramzadeh50–Reference Omrani, Rahimi and Nikseresht52,Reference Atapour, Vahdat and Hosseini55,Reference Assarzadeh, Vahdat and Seirafian56) . Six studies showed a high risk of bias in this domain because they did not show enough information about the losses and their reasons(Reference Napolitano, Bonomini and Bomba45,Reference Temple, Smith and Cockram46,Reference Zachara, Gromadzinska and Zbrog48,Reference Zachara, Gromadzinska and Palus49,Reference Salimian, Soleimani and Bahmani53,Reference Jamal, Seifati and Dehghani Ashkezari54) .
Only four studies had some concerns in domain measurement of the outcome for not presenting information about the primary outcome(Reference Salehi, Sohrabi and Ekramzadeh50,Reference Salimian, Soleimani and Bahmani53,Reference Jamal, Seifati and Dehghani Ashkezari54,Reference Assarzadeh, Vahdat and Seirafian56) . All the studies had some concerns in domain five (selection of the reported results) because they did not give enough information, such as the analysis plan or the primary outcome.
Plasma selenium
Six studies compared the selenium status between HD patients and healthy subjects at baseline(Reference Bonomini, Forster and De Risio44–Reference Zachara, Gromadzinska and Palus49). In five of these studies, HD patients had lower plasma Se levels compared with healthy individuals(Reference Bonomini, Forster and De Risio44,Reference Napolitano, Bonomini and Bomba45,Reference Zachara, Adamowicz and Trafikowska47–Reference Zachara, Gromadzinska and Palus49) . Other three studies reported that HD patients were considered Se deficient when compared with the healthy range for plasma Se concentration(Reference Omrani, Golmohamadi and Pasdar51,Reference Omrani, Rahimi and Nikseresht52,Reference Atapour, Vahdat and Hosseini55) . Only one study showed that patients had plasma Se levels with no difference from healthy individuals(Reference Temple, Smith and Cockram46), and four studies did not report information on Se status at the baseline(Reference Salehi, Sohrabi and Ekramzadeh50,Reference Salimian, Soleimani and Bahmani53,Reference Jamal, Seifati and Dehghani Ashkezari54,Reference Assarzadeh, Vahdat and Seirafian56) .
Nine studies found higher Se levels in HD patients compared with the control group of patients and/or baseline data after the supplementation (range of doses: 140 ± 9 µg/d to 500 µg/d; range of time: 14 d to 6 months)(Reference Bonomini, Forster and De Risio44–Reference Zachara, Gromadzinska and Palus49,Reference Omrani, Golmohamadi and Pasdar51,Reference Omrani, Rahimi and Nikseresht52,Reference Atapour, Vahdat and Hosseini55) . Four studies did not report data on plasma selenium(Reference Salehi, Sohrabi and Ekramzadeh50,Reference Salimian, Soleimani and Bahmani53,Reference Jamal, Seifati and Dehghani Ashkezari54,Reference Assarzadeh, Vahdat and Seirafian56) (Table 2).
Thyroid function
Two studies assessed the effects of Se supplementation on thyroid hormones. Napolitano et al. (1996)(Reference Napolitano, Bonomini and Bomba45), who provided Se supplementation three times a week with doses of 500 µg/d in the first 3 months and 200 µg/d in the last 3 months, showed that, at the beginning of the study, the patients had thyroid function in the normal range. After 2 months of supplementation, a substantial increase in free triiodothyronine (FT3) compared with baseline was observed (2.72 ± 0.32 and 2.38 ± 0.31 pg/ml, respectively); however, this was not the case at 6 months. After 6 months, a reduction in thyroid-stimulating hormone (TSH) was observed in the supplemented group compared with baseline data (1.19 ± 0.38 and 2.13 ± 0.85 mIU/l, respectively). Omrani et al. (2015)(Reference Omrani, Golmohamadi and Pasdar51) demonstrated that the group supplemented for 3 months with 200 µg/d of Se did not show any alteration compared with the control group in TSH (3.7 ± 2.22 and 2.84 ± 1.88 µU/ml, respectively), free thyroxine (FT4) (7.19 ± 1.98 and 7.02 ± 1.87 µg/dl, respectively) and T3 resin uptake (T3RU) (30.04 ± 2.28% and 29.2 ± 1.98%, respectively) (Table 2).
Immunological parameters
Only one study analysed immunological parameters in response to Se supplementation(Reference Bonomini, Forster and De Risio44). The patients were supplemented with Se three times a week for 6 months, with 500 µg/d in the first 3 months and 200 µg/d in the following 3 months in two different cities (Chieti and Rostock). After 6 months of supplementation, no change was observed in white blood counts (Chieti: 6060 ± 1957 and Rostock: 7280 ± 3325), total lymphocytes (Chieti: 1569 ± 413 and Rostock: 1840 ± 1179) and lymphocyte subpopulations (CD2: Chieti: 85 ± 4% and Rostock: 83 ± 3%; CD3: Chieti: 58 ± 11% and Rostock: 74 ± 2%; CD4: Chieti: 42 ± 6% and Rostock: 45 ± 6%; CD8: Chieti: 32 ± 7% and Rostock: 37 ± 10%; CD20: Chieti: 6 ± 1% and Rostock: 17 ± 2%; CD4+ CD45R+: Chieti: 33 ± 7% and Rostock: 30 ±14%), which were within the normal range, compared with pre-supplementation data and with the control group. However, an improvement was observed in T-cell response (94 252 ± 33 526 versus 118 944 ± 18 642 counts per minute (c.p.m.)) and an increase in delayed-type hypersensitivity compared with the control group (Table 2).
Oxidative stress and inflammatory markers
A total of ten studies assessed the effects of Se supplementation on oxidative stress or inflammatory markers(Reference Bonomini, Forster and De Risio44,Reference Temple, Smith and Cockram46–Reference Omrani, Golmohamadi and Pasdar51,Reference Salimian, Soleimani and Bahmani53,Reference Jamal, Seifati and Dehghani Ashkezari54,Reference Assarzadeh, Vahdat and Seirafian56) . Bonomini et al. (1995)(Reference Bonomini, Forster and De Risio44) showed an increase in whole blood GPx activity (11 005.6 ± 3599.1 IU/l) compared with the control group (6673.2 ± 1107.4 IU/l) after Se supplementation for 6 months (500 µg/d in the first 3 months and 200 µg/d of Se in the last 3 months, three times a week). The most change was found after 4 months of supplementation and then stabilised. On the other hand, the study by Temple et al. (2000)(Reference Temple, Smith and Cockram46) performed for 14 d with 200 μg/d of Se demonstrated that, before supplementation, GPx activities in HD patients were within the normal range of healthy subjects. After supplementation, there was no difference between the supplemented and control groups regarding plasma (selenite: 1.74 ± 0.11 and selenate: 1.76 ± 0.14 versus 7.82 ± 0.27 U/g protein, respectively) and erythrocyte (selenite: 31.6 ± 21 and selenate: 33.3 ± 2.7 versus 31.4 ± 2.1 U/g Hemoglobin (Hgb), respectively) GPx activity (Table 2).
The study by Zachara et al. (2001)(Reference Zachara, Adamowicz and Trafikowska47) found that HD patients had lower GPx activity (as measured in plasma (102 U/l) and erythrocytes (13.1 U/g Hemoglobin (Hb)) in comparison with healthy subjects (224 U/l; and 19.1 U/g Hb, respectively), and higher erythrocyte reduced glutathione (GSH) concentration (2.77 mmol/l) compared with healthy subjects at baseline (2.32 mmol/l). After Se supplementation for 3 months with 200 μg/d of Se, erythrocyte GPx activity increased from baseline, reaching a plateau equated to healthy subjects, while plasma GPx activity and erythrocyte GSH did not change. In another study by Zachara et al. (2009)(Reference Zachara, Gromadzinska and Zbrog48), HD patients had lower plasma GPx levels (11.4 ± 6.7 µg/ml) than healthy individuals (48.4 ± 12.3 µg/ml) at baseline. However, no changes in plasma GPx were identified among the non-supplemented and supplemented group of patients after 3 months of 200 μg/d of Se supplementation. Further, Salimian et al. (2022)(Reference Salimian, Soleimani and Bahmani53) showed an increase in total GSH (366.4 ± 60.8 to 424.3 ± 48.4; effect size 0.61) in HD patients supplementing Se (200 μg/d) for 24 weeks (Table 2).
Salehi et al. (2013)(Reference Salehi, Sohrabi and Ekramzadeh50) reported a decrease in malondialdehyde (MDA), which is a lipid peroxidation marker, in the supplemented group (200 µg Se/d for 12 weeks) (−1.2 (−3.17, −0.45) µmol/l (confidence interval 95% (CI) −2.41, −1.16)) while the non-supplemented patients presented a rise in the same period (0.6 (0.01, 2.22) µmol/l (CI 0.47, 1.86)) and no change in the high-sensitivity C-reactive protein (hs-CRP) (−0.85 (−2.47, 5.25) μg/ml (CI−1.99, 16.46)) compared with the control group (1.3 (−17.7, 4.52) μg/ml (CI−13.69, 5.08)). Assarzadeh et al. (2023)(Reference Assarzadeh, Vahdat and Seirafian56), who supplemented 200 µg Se/d for 3 months, also showed no change in hs-CRP between the groups (intervention group: 0.24 ± 0.5 mg/l and control group: 0.6 ± 0.8 mg/l). Omrani et al. (2015)(Reference Omrani, Golmohamadi and Pasdar51) showed no difference in the CRP concentration between groups (supplemented group: 14.77 ± 17.93 and control group: 18.29 ± 21.56) after supplementation of 200 µg/d of Se for 3 months). On the other hand, a decrease in CRP was observed in the supplemented group (4.5 ± 1.7 to 3.9 ± 1.4 mg/l; effect size 0.37) (200 μg/d of Se for 24 weeks) by Salimian et al. (2022)(Reference Salimian, Soleimani and Bahmani53), while no differences in MDA were seen (2.8 ± 0.3 to 2.6 ± 0.2 µmol/l; effect size 0.02) (Table 2).
Salehi et al. (2013)(Reference Salehi, Sohrabi and Ekramzadeh50) also reported an increase in serum interleukin-6 (IL-6) levels in control (22.95 (0.92, 1978.2) pg/ml (CI 336.33, 1218.3)) and supplemented groups (6.05 (−20.4, 50.8) pg/ml (CI −61.4, 89.68)) after 200 µg Se/d for 12 weeks; however, Jamal et al. (2022)(Reference Jamal, Seifati and Dehghani Ashkezari54) supplemented 200 μg/d of Se for 24 weeks and observed a decrease in interleukin-1 (IL-1) and tumour necrosis factor-α (TNF-α), no changes in interleukin-8 (IL-8) and an increase in the expression of peroxisome proliferator-activated receptor-gamma (PPAR-γ), and transforming growth factor-β (TGF-β) in the treated group (Table 2).
The study by Zachara et al. (2011)(Reference Zachara, Gromadzinska and Palus49) showed that, at baseline, HD patients had higher levels of DNA damage in leucocytes (0.73 ± 0.84) and oxidative base lesions in DNA (1.28 ± 1.60) compared with healthy subjects (0.25 ± 0.24 and 0.55 ± 0.45, respectively). After supplementation of 200 μg/d of Se for 3 months, the DNA damage was reduced in the supplemented group compared with baseline data, almost two times lower compared with the placebo group. In addition, oxidative base lesions in DNA were also reduced (1.12 ± 1.01 to 0.43 ± 0.52) compared with baseline data and the non-supplemented control group (Table 2).
Discussion
This is the first systematic review to comprehensively analyse the results of clinical trials on Se supplementation in patients with CKD undergoing HD. The majority of studies reported that HD patients have low Se status, evidenced by lower plasma Se concentration (baseline range 32–51 μg/l) compared with healthy individuals or the expected healthy range. Se supplementation improved the Se status of HD patients, shifting their plasma Se concentrations (post-treatment range 66.65–182 µg/l)(Reference Bonomini, Forster and De Risio44–Reference Zachara, Gromadzinska and Palus49,Reference Omrani, Golmohamadi and Pasdar51,Reference Omrani, Rahimi and Nikseresht52,Reference Atapour, Vahdat and Hosseini55) into the normal range of 60–120 µg/l(Reference Van Dael and Deelstra57,Reference Ortuño, Ros and Periago58) .
Our findings corroborate previous research that reported that patients with CKD undergoing HD are more vulnerable to micronutrient deficiency due to loss of appetite and accumulation of uremic toxins, loss of trace elements during the HD procedure, and impaired intestinal absorption due to gut dysbiosis and disruption of the intestinal epithelial barrier(Reference Gómez de Oña, Martínez-Morillo and Gago González18,Reference Bossola, Di Stasio and Viola59–Reference Zachara, Gromadzińska and Wasowicz61) .
Despite studies showing that Se supplementation raised plasma Se levels, they did not thoroughly evaluate many other factors in the secondary outcomes. Furthermore, the studies that did look forward to secondary outcome parameters had conflicting results, and there were not enough studies to provide a consistent conclusion.
Adequate Se status is essential for thyroid function as there are selenoproteins involved in the metabolism of its hormones and in maintaining the gland integrity. These include GPx and thioredoxin TrxR, which are involved with oxidative stress processes, and DIO, involved in activating or inactivating thyroid hormones through its ability to catalyse deiodination, acts in the conversion of thyroxine (T4) to triiodothyronine (T3) and T4 to RT3(Reference Papp, Lu and Holmgren4,Reference Iglesias, Selgas and Romero62,Reference Köhrle63) . However, of the two studies included that analysed thyroid hormones, only one showed positive results after supplementation with a reduction in TSH and an increase in FT3(Reference Napolitano, Bonomini and Bomba45). According to Thomson et al. (2009)(Reference Thomson, Campbell and Miller64), in cases of Se deficiency, its levels are maintained in the thyroid, and the deiodinase enzymes are at the top of the selenoprotein hierarchy, so insufficient Se levels are required to modify the thyroid functions. This suggests that plasma levels of Se might not be reduced enough to verify changes in thyroid function. Se deficiency may play a role in thyroid dysfunction; it is worth considering that, in the CKD context, these patients commonly present thyroid alterations and disorders associated with the uraemic environment. This can cause an impaired response of the pituitary gland to thyrotropin-releasing hormone (TRH), resulting in impaired clearance and prolonged half-life of TSH. Additionally, inflammatory cytokines and metabolic acidosis can affect the synthesis and conversion of T4 and T3 hormones. Patients undergoing dialysis may also experience significant protein loss, which can further contribute to low levels of T3 and T4 hormones(Reference Iglesias and Díez65–Reference Mohamedali, Reddy Maddika and Vyas68).
Because of its antioxidant function, Se has the potential to impact the immune system. Selenoproteins can modulate and facilitate immune responses; for example, selenoprotein K is essential for T-cell proliferation, neutrophil migration and cell protection from apoptosis induced by oxidative stress(Reference Papp, Lu and Holmgren4,Reference Ye, Huang and Wang69) . The thymus, a gland with Iodothyronine deiodinase 2 (DIO2) activity, may experience altered cell function and negative effects on the immune system due to reduced Se levels(Reference Ye, Huang and Wang69,Reference Arthur, McKenzie and Beckett70) . In addition, a decrease in Se concentration can generate a lower expression of GPx, affecting the inflammatory signalling capacity of macrophages and the ability of lymphocytes to proliferate, since the lymphocytes are sensitive to oxidative stress(Reference Avery and Hoffmann1,Reference Ye, Huang and Wang69,Reference Arthur, McKenzie and Beckett70) . Despite the critical role of Se in the immune system of HD patients, only one study included in this review analysed immunological parameters and showed improvement in T-cell response(Reference Bonomini, Forster and De Risio44).
Se supplementation can increase the expression of the selenoprotein genes and, consequently, inhibit nuclear factor kappa-B (NF-kB), which is involved in the expression of inflammatory cytokines(Reference Duntas71). The GPx family represents one of the most important antioxidant enzymes that act against oxidative stress, reducing the accumulation of reactive oxygen species (ROS), and working against inflammatory processes related to oxidative reactions,(Reference Zachara, Gromadzińska and Wasowicz61,Reference Iglesias, Selgas and Romero62,Reference Godos, Giampieri and Micek72) . The kidneys are identified as the main source of plasma GPx synthesis, while the erythrocyte GPx seems to be unrelated to renal function(Reference Avissar, Ornt and Yagil33). Therefore, a damaged kidney tissue as in patients with CKD could lead to its inability to synthesise plasma GPx, as demonstrated by the study of Schiavon et al. (1994)(Reference Schiavon, Guidi and Biasioli73). The authors showed that a decrease in GPx levels is common in most renal diseases and is correlated with indicators of kidney function. These results justify the GPx non-alteration in some studies, despite Se supplementation. Additionally, in studies that have shown positive results, a compensatory response to the higher production of ROS in HD patients may be suggested, as they have lower activities of antioxidant enzymes(Reference Zachara, Adamowicz and Trafikowska47).
The GPx enzyme also plays an essential role against oxidative DNA damage caused by ROS reaction at the base guanine, leading to nucleobase oxidation and DNA damage, forming as the product 8-hydroxy-2-deoxyguanosine (8-OHdG)(Reference Sung, Hsu and Chen74–Reference López-Uriarte, Nogués and Saez77). The selenoproteins also protect against oxidative damage in the lipids in cell membranes, and in cases of Se deficiency, lipid peroxidation may occur, leading to the formation of the final product MDA and an increased inflammatory process(Reference Avery and Hoffmann1,Reference Godos, Giampieri and Micek72) . Only the study by Salehi et al. (2013)(Reference Salehi, Sohrabi and Ekramzadeh50) found a reduction in IL-6 and MDA levels.
Jamal et al. (2022)(Reference Jamal, Seifati and Dehghani Ashkezari54) showed that Se supplementation can improve inflammatory processes by decreasing IL-1 and TNF-α and increasing the expression of PPAR-γ and TGF-β. This was the only study that analysed these parameters, and the inhibition of NF-kB could explain these results. In addition, the study by Salimian et al. (2022)(Reference Salimian, Soleimani and Bahmani53) shows a decrease in CRP, representing a reduction in inflammation in the patients; however, the results regarding inflammation parameters seem to be conflicted with the other studies.
Factors such as duration of supplementation, doses administered and type of supplementation should be considered when verifying differences in secondary outcomes among the studies. Studies that used higher doses or supplemented for a more extended period showed some positive changes(Reference Bonomini, Forster and De Risio44,Reference Napolitano, Bonomini and Bomba45,Reference Zachara, Adamowicz and Trafikowska47,Reference Salehi, Sohrabi and Ekramzadeh50,Reference Salimian, Soleimani and Bahmani53,Reference Jamal, Seifati and Dehghani Ashkezari54) , as well as studies that used the organic form of Se as the Se-rich yeast(Reference Zachara, Adamowicz and Trafikowska47,Reference Zachara, Gromadzinska and Palus49,Reference Jamal, Seifati and Dehghani Ashkezari54) , a form of supplementation with a predominant concentration of selenomethionine which is characterised by greater bioavailability than inorganic forms(Reference Klopotek, Hirche and Eder78). Even so, not all parameters analysed in each of the studies showed positive changes.
As these results show no consensus on the doses that should be used and the time of supplementation for a positive effect of Se, further studies should be carried out. In addition, only one of the studies presented an analysis of the HD patients’ food intake to verify the dietary intake of Se(Reference Salimian, Soleimani and Bahmani53). It is crucial to verify the Se needs of these patients, as their adequate levels to demonstrate positive results of their functions may be different from the parameters established for a healthy population. Furthermore, it becomes essential to define at what moment Se supplementation is necessary so that positive changes can be maximised.
It is important to note that three studies used amounts of Se ranging from 400 to 500 µg/d(Reference Bonomini, Forster and De Risio44,Reference Napolitano, Bonomini and Bomba45,Reference Atapour, Vahdat and Hosseini55) , which include doses on the limit or above the UL. Higher doses of Se can lead to adverse health effects such as alopecia, dermatitis, neuropsychological changes and an increased risk of mortality(Reference Rayman13,Reference Turck and Bohn36) . Recent studies indicate that exposure to Se at concentrations considered healthy (below the UL)(38) is associated with an increased risk of diabetes, non-alcoholic liver disease and insulin resistance(Reference Vinceti, Filippini and Rothman37,Reference Yang, Yan and Liu79,Reference Cardoso, Braat and Graham80) . Further, a recent recommendation from the European Food Safety Authority suggested decreasing the UL to 255 μg/d(Reference Turck and Bohn36). The recommended safe intake range for Se is quite narrow and is still a topic of debate. Therefore, Se supplementation should only be considered depending on the objective, associated risk and losses of Se in biological fluids and for populations with low levels of Se(Reference Berger, Shenkin and Schweinlin32,Reference Turck and Bohn36,Reference Cardoso, Braat and Graham80,Reference Vinceti, Filippini and Del Giovane81) . As seen previously, most of the studies reported that HD patients have low Se status which could be related to the loss during the HD process, inflammation, redistribution of circulating selenium, and oxidative stress(Reference Bossola, Di Stasio and Viola59,Reference Tonelli, Wiebe and Bello60,Reference Canaud, Cristol and Morena82,Reference Rucker, Thadhani and Tonelli83) , but there are still no specific dietary recommendations of Se for these patients, who currently follow recommendations meant for the general population(Reference Ikizler, Burrowes and Byham-Gray40).
The present review has some limitations that must be acknowledged. First, all the studies included here presented a high risk of bias, making it difficult to reach a satisfactory conclusion on the effects of Se supplementation, so more studies with improved study designs are necessary, with an emphasis on the randomisation process, appropriate analysis of the intended interventions, development of a clinical trial flowchart, and informed prior planning of the analyses that were carried out. In addition, due to the heterogeneity and low quality of the included studies, it was not possible to conduct a meta-analysis. Furthermore, some studies also presented limited information affecting the results’ reliability. For instance, Omrani et al. (2016)(Reference Omrani, Rahimi and Nikseresht52) did not provide the dose of Se supplementation or the unit of measurement of Se, and Salehi et al. (2013)(Reference Salehi, Sohrabi and Ekramzadeh50) did not report plasma Se levels before and after supplementation. Also, most studies did not report disease aetiology and clinical outcomes data, such as mortality. Although not the objective of this review, this information could contribute to this study findings. Further, Bonomini et al. (1995)(Reference Bonomini, Forster and De Risio44) and Napolitano et al. (1996)(Reference Napolitano, Bonomini and Bomba45) used doses above the UL during the first 3 months of supplementation, doses that are associated with adverse health effects. Despite these limitations, this review has strengths, as it is the first systematic review to address this topic; the search strategy was elaborated with the help of a professional experienced in systematic review so that the search for articles could be performed more sensitively and all the recommendations and criteria established for conducting a systematic review were performed.
Conclusion
The results of this review suggest that Se supplementation, with a dose of approximately 200 µg/d, can increase plasma Se levels in patients with CKD undergoing HD. However, evidence is limited in demonstrating the beneficial effects of Se supplementation on antioxidant and inflammatory markers, immune response and thyroid hormones. Given that all the included studies in this review were classified as having a high risk of bias, we highlight the need for further studies with an improvement of the study designs and data reports to elucidate and explore in more detail the effects of Se supplementation in patients with CKD undergoing HD.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0954422424000222
Acknowledgements
We would like to thank Daniele MTP Ferreira for her contribution to elaborating the search strategies to identify the scientific material.
Financial support
This study was supported by Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ) (Process E-26/201.373/2021 and E-26/200.747/2021).
Competing interests
The authors have no conflicts of interest to declare.
Authorship
The authors’ responsibilities were as follows: P.C.T., M.B.S.P. and M.C.A. were responsible for the designed research; P.C.T., V.O.L. and M.B.S.P. conducted research and performed the data analysis; P.C.T. was responsible for writing the manuscript; M.B.S.P., B.R.C., D.M., M.C.A. and V.O.L. performed a critical review of the manuscript. All authors read and approved the final version manuscript.
