Scientific opinion on the tolerable upper intake level for selenium

. 2023 Jan 20;21(1):e07704. doi: 10.2903/j.efsa.2023.7704

Biomarker	Description	Influencing factors	Characteristics	Sensitivity as biomarker of selenium intake^(b)
Serum/plasma ^(a) Se concentrations	Non‐cellular Se; organic (mainly selenoproteins (selenoprotein P, glutathione peroxidase 3, thioredoxin reductases)); albumin‐bound SeMet; selenosugars (Ward‐Deitrich et al., 2021) and inorganic selenium (Vinceti et al., 2015a; Filippini et al., 2018)	‐ Sex, age (Sheehan and Halls, 1999; Muntau et al., 2002) ‐ Smoking status (Swanson et al., 1990; Alfthan and Neve, 1996; Kocyigit et al., 2001; Arnund et al., 2006; Filippini et al., 2018) ‐ Inflammation, disease (Miller et al., 1983; Nichol et al., 1998; Combs et al., 2012) ‐ Chemical nature of dietary Se ‐ GPX1 679 genotype (Combs et al., 2012)	‐ Varies across geographical locations (Fordyce, 2013; Stoffaneller and Morse, 2015; Ibrahim et al., 2019) ‐ Responds to Se supplementation over wide range of supplemental intake (15–700 μg/d); dose‐dependent (Ashton et al., 2009) ‐ Higher response to organic Se, especially in non‐deficient populations (Thomson et al., 1982; Burk et al., 2006; Combs, 2015) ‐ Suitable for speciation analysis	‐ Can detect changes in intake over short/medium‐term ‐ Can distinguish ‘high’ from ‘low’ consumers ‐ Low responsiveness to inorganic Se intake in Se‐replete populations (Neve, 1995) ‐ Population‐specific equations to predict dietary Se intake (Longnecker et al., 1996; Burk et al., 2006; Combs, 2015)
RBC Se concentrations	Mainly glutathione peroxidase 1 and Hb‐bound Se	‐ Age (Lloyd et al., 1983) ‐ Chemical nature of dietary Se (Neve, 1995) ‐ Unaffected by inflammatory responses (Stefanowicz et al., 2013)	‐ Responds to Se supplementation, more slowly than serum/plasma (Neve, 1995; Ashton et al., 2009)	‐ Can detect changes in intake over medium/long‐term ‐ Can distinguish ‘high’ from ‘low’ consumers
Whole blood Se concentrations	Cellular and non‐cellular circulating Se	‐ Age (Lloyd et al., 1983) ‐ Smoking status (Lloyd et al., 1983) ‐ Disease (Muecke et al., 2009; Muecke et al., 2018) ‐ Chemical nature of dietary Se	‐ Varies across geographical locations (Fordyce, 2013) ‐ Responds to Se supplementation over wide range of supplemental intake (15–200^(c) μg/d); dose‐dependent (Neve, 1995; Ashton et al., 2009) ‐ Suitable for speciation analysis	‐ Can detect changes in intake over medium/long‐term ‐ Can distinguish ‘high’ from ‘low’ consumers ‐ Population‐specific to predict dietary Se intake (Yang et al., 1989b)
Urine Se concentrations	Primary route of Se elimination, mainly as selenosugars 1 and 3, trimethylselenonium ion (TMSe) and selenate and Se‐methylselenoneine	‐ Sex ‐ Chemical nature of dietary Se (Burk et al., 2006) ‐ Se status ‐ Kidney function (Oster and Prellwitz, 1990) ‐ Physical activity (Rodriguez Rodriguez et al., 1995) ‐ GPX1 679 genotype (Combs et al., 2012) ‐ TMSe eliminators vs non‐eliminators (Kuehnelt et al., 2015; Lajin et al., 2016a)	‐ Varies across geographical locations (Fordyce, 2013) ‐ Responds to Se supplementation over wide range of supplemental intake (100–700 μg/d); dose‐dependent (Yang et al., 1989b; Burk et al., 2006; Combs et al., 2012) ‐ Correlates with plasma Se over a wide range (Yang et al., 1989b; Sanz Alaejos and Diaz Romero, 1993) ‐ Suitable for speciation analysis	‐ Can detect changes in intake over short term ‐ Can distinguish ‘high’ from ‘low consumers’ ‐ Population‐specific to predict dietary Se intake (Yang et al., 1989b; Sanz Alaejos and Diaz Romero, 1993; Longnecker et al., 1996; Combs, 2015)
(Toe)nail/hair Se concentrations	Deposition of Se	‐ Age (Hunter et al., 1990) ‐ Smoking (Swanson et al., 1990; van den Brandt et al., 1993; Virtanen et al., 1996; Xun et al., 2011) ‐ Chemical nature of dietary Se (i.e. toenail Se content might not reflect inorganic Se exposure (Filippini et al., 2017)) ‐ Growth rate (Slotnick and Nriagu, 2006) ‐ External exposure (e.g. Se‐containing shampoos)	‐ Varies across geographical locations (Morris et al., 1983; Hunter et al., 1990; Fordyce, 2013) ‐ Responds slowly to Se supplementation (weeks to months) (Gallagher et al., 1984; Longnecker et al., 1993) ‐ Correlation with Se intake (Hunter et al., 1990; Swanson et al., 1990) and blood/plasma Se (Yang et al., 1989a; Satia et al., 2006) over a wide range shown in some studies. Other studies found no or low association with Se intake (Hunter et al., 1990; Al‐Saleh and Billedo, 2006; Satia et al., 2006; Vinceti et al., 2012) or blood selenium (Satia et al., 2006; Vinceti et al., 2012; Chawla et al., 2020).	‐ Can detect changes in intake over medium/long‐term ‐ Can distinguish ‘high’ from ‘low’ consumers ‐ Requires standardised procedures for sample collections and treatments as prone to contamination (Slotnick and Nriagu, 2006) ‐ Population‐specific to predict dietary Se intake based on toenail content (Longnecker et al., 1996)
Plasma selenoprotein P concentrations	20%–70% of total plasma Se; mostly secreted in the liver (Saito, 2021)	‐ Inflammation; oxidative stress (Saito, 2020; Vinceti et al., 2022); insulin levels and glucose metabolism (Speckmann et al., 2008; Mao and Teng, 2013; Schomburg, 2022) ‐ Se status ‐ Selenoprotein P polymorphisms (Meplan et al., 2007)	‐ Responds rapidly to Se supplementation in populations with ‘low Se’ intake/status (Duffield et al., 1999; Hurst et al., 2010) ‐ Correlates with plasma/serum Se up to 80–90 μg/L (Hurst et al., 2013)	‐ Responsive in population with Se status in the lowest range ‐ Plateau (i.e. maximum expression) in the higher range of Se intake (> ca. 60–70 μg/day)
Plasma, RBC, platelet and whole blood glutathione peroxidase activity	‐ Glutathione peroxidase 3 in plasma (represents 10%–25% of plasma/whole blood Se) ‐ Glutathione peroxidase 1 in RBC ‐ Glutathione peroxidase 1 and glutathione peroxidase 4 in platelets	‐ Sex, age ‐ Se status ‐ Race ‐ Physical activity (Tessier et al., 1995; Pograjc et al., 2012) ‐ Deficiencies in other nutrients ‐ Chemical nature of dietary Se (Thomson et al., 1982; Xia et al., 2005) ‐ Diseases, polymorphisms (Hurst et al., 2013)	‐ Plasma glutathione peroxidase is a useful marker of Se intake/status in populations with low Se intake; it responds rapidly to supplementation. Correlates with plasma/serum Se up to 80 μg/L serum Se (Rea et al., 1979; Muller et al., 2020) ‐ RBC GPx glutathione peroxidase responds slowly to depletion and supplementation; plateau at plasma Se levels > 100 μg/L (Neve, 1995; zzn et al., 2000; Burk et al., 2006; Hurst et al., 2010; Combs et al., 2012) ‐ Platelet glutathione peroxidase responds rapidly to Se dietary changes; plateau at plasma Se levels > 100 μg/L (Alfthan et al., 1991; Neve, 1995; Burk et al., 2006; Hurst et al., 2010)	‐ Responsive in population with Se status in the lowest range ‐ Plateau in the higher range of Se intakes (> ca. 40–60 μg/day)

Biomarker

Description

Influencing factors

Characteristics

Sensitivity as biomarker of selenium intake^(b)

Serum/plasma ^(a) Se concentrations

Non‐cellular Se; organic (mainly selenoproteins (selenoprotein P, glutathione peroxidase 3, thioredoxin reductases)); albumin‐bound SeMet; selenosugars (Ward‐Deitrich et al., 2021) and inorganic selenium (Vinceti et al., 2015a; Filippini et al., 2018)

‐ Sex, age (Sheehan and Halls, 1999; Muntau et al., 2002)

‐ Smoking status (Swanson et al., 1990; Alfthan and Neve, 1996; Kocyigit et al., 2001; Arnund et al., 2006; Filippini et al., 2018)

‐ Inflammation, disease (Miller et al., 1983; Nichol et al., 1998; Combs et al., 2012)

‐ Chemical nature of dietary Se

‐ GPX1 679 genotype (Combs et al., 2012)

‐ Varies across geographical locations (Fordyce, 2013; Stoffaneller and Morse, 2015; Ibrahim et al., 2019)

‐ Responds to Se supplementation over wide range of supplemental intake (15–700 μg/d); dose‐dependent (Ashton et al., 2009)

‐ Higher response to organic Se, especially in non‐deficient populations (Thomson et al., 1982; Burk et al., 2006; Combs, 2015)

‐ Suitable for speciation analysis

‐ Can detect changes in intake over short/medium‐term

‐ Can distinguish ‘high’ from ‘low’ consumers

‐ Low responsiveness to inorganic Se intake in Se‐replete populations (Neve, 1995)

‐ Population‐specific equations to predict dietary Se intake (Longnecker et al., 1996; Burk et al., 2006; Combs, 2015)

RBC Se concentrations

Mainly glutathione peroxidase 1 and Hb‐bound Se

‐ Age (Lloyd et al., 1983)

‐ Chemical nature of dietary Se (Neve, 1995)

‐ Unaffected by inflammatory responses (Stefanowicz et al., 2013)

‐ Responds to Se supplementation, more slowly than serum/plasma (Neve, 1995; Ashton et al., 2009)

‐ Can detect changes in intake over medium/long‐term

‐ Can distinguish ‘high’ from ‘low’ consumers

Whole blood Se concentrations

Cellular and non‐cellular circulating Se

‐ Age (Lloyd et al., 1983)

‐ Smoking status (Lloyd et al., 1983)

‐ Disease (Muecke et al., 2009; Muecke et al., 2018)

‐ Chemical nature of dietary Se

‐ Varies across geographical locations (Fordyce, 2013)

‐ Responds to Se supplementation over wide range of supplemental intake (15–200^(c) μg/d); dose‐dependent (Neve, 1995; Ashton et al., 2009)

‐ Suitable for speciation analysis

‐ Can detect changes in intake over medium/long‐term

‐ Can distinguish ‘high’ from ‘low’ consumers

‐ Population‐specific to predict dietary Se intake (Yang et al., 1989b)

Urine Se concentrations

Primary route of Se elimination, mainly as selenosugars 1 and 3, trimethylselenonium ion (TMSe) and selenate and Se‐methylselenoneine

‐ Sex

‐ Chemical nature of dietary Se (Burk et al., 2006)

‐ Se status

‐ Kidney function (Oster and Prellwitz, 1990)

‐ Physical activity (Rodriguez Rodriguez et al., 1995)

‐ GPX1 679 genotype (Combs et al., 2012)

‐ TMSe eliminators vs non‐eliminators (Kuehnelt et al., 2015; Lajin et al., 2016a)

‐ Varies across geographical locations (Fordyce, 2013)

‐ Responds to Se supplementation over wide range of supplemental intake (100–700 μg/d); dose‐dependent (Yang et al., 1989b; Burk et al., 2006; Combs et al., 2012)

‐ Correlates with plasma Se over a wide range (Yang et al., 1989b; Sanz Alaejos and Diaz Romero, 1993)

‐ Suitable for speciation analysis

‐ Can detect changes in intake over short term

‐ Can distinguish ‘high’ from ‘low consumers’

‐ Population‐specific to predict dietary Se intake (Yang et al., 1989b; Sanz Alaejos and Diaz Romero, 1993; Longnecker et al., 1996; Combs, 2015)

(Toe)nail/hair Se

concentrations

Deposition of Se

‐ Age (Hunter et al., 1990)

‐ Smoking (Swanson et al., 1990; van den Brandt et al., 1993; Virtanen et al., 1996; Xun et al., 2011)

‐ Chemical nature of dietary Se (i.e. toenail Se content might not reflect inorganic Se exposure (Filippini et al., 2017))

‐ Growth rate (Slotnick and Nriagu, 2006)

‐ External exposure (e.g. Se‐containing shampoos)

‐ Varies across geographical locations (Morris et al., 1983; Hunter et al., 1990; Fordyce, 2013)

‐ Responds slowly to Se supplementation (weeks to months) (Gallagher et al., 1984; Longnecker et al., 1993)

‐ Correlation with Se intake (Hunter et al., 1990; Swanson et al., 1990) and blood/plasma Se (Yang et al., 1989a; Satia et al., 2006) over a wide range shown in some studies. Other studies found no or low association with Se intake (Hunter et al., 1990; Al‐Saleh and Billedo, 2006; Satia et al., 2006; Vinceti et al., 2012) or blood selenium (Satia et al., 2006; Vinceti et al., 2012; Chawla et al., 2020).

‐ Can detect changes in intake over medium/long‐term

‐ Can distinguish ‘high’ from ‘low’ consumers

‐ Requires standardised procedures for sample collections and treatments as prone to contamination (Slotnick and Nriagu, 2006)

‐ Population‐specific to predict dietary Se intake based on toenail content (Longnecker et al., 1996)

Plasma selenoprotein P concentrations

20%–70% of total plasma Se; mostly secreted in the liver (Saito, 2021)

‐ Inflammation; oxidative stress (Saito, 2020; Vinceti et al., 2022); insulin levels and glucose metabolism (Speckmann et al., 2008; Mao and Teng, 2013; Schomburg, 2022)

‐ Se status

‐ Selenoprotein P polymorphisms (Meplan et al., 2007)

‐ Responds rapidly to Se supplementation in populations with ‘low Se’ intake/status (Duffield et al., 1999; Hurst et al., 2010)

‐ Correlates with plasma/serum Se up to 80–90 μg/L (Hurst et al., 2013)

‐ Responsive in population with Se status in the lowest range

‐ Plateau (i.e. maximum expression) in the higher range of Se intake (> ca. 60–70 μg/day)

Plasma, RBC, platelet and whole blood glutathione peroxidase activity

‐ Glutathione peroxidase 3 in plasma (represents 10%–25% of plasma/whole blood Se)

‐ Glutathione peroxidase 1 in RBC

‐ Glutathione peroxidase 1 and glutathione peroxidase 4 in platelets

‐ Sex, age

‐ Se status

‐ Race

‐ Physical activity (Tessier et al., 1995; Pograjc et al., 2012)

‐ Deficiencies in other nutrients

‐ Chemical nature of dietary Se (Thomson et al., 1982; Xia et al., 2005)

‐ Diseases, polymorphisms (Hurst et al., 2013)

‐ Plasma glutathione peroxidase is a useful marker of Se intake/status in populations with low Se intake; it responds rapidly to supplementation. Correlates with plasma/serum Se up to 80 μg/L serum Se (Rea et al., 1979; Muller et al., 2020)

‐ RBC GPx glutathione peroxidase responds slowly to depletion and supplementation; plateau at plasma Se levels > 100 μg/L (Neve, 1995; zzn et al., 2000; Burk et al., 2006; Hurst et al., 2010; Combs et al., 2012)

‐ Platelet glutathione peroxidase responds rapidly to Se dietary changes; plateau at plasma Se levels > 100 μg/L (Alfthan et al., 1991; Neve, 1995; Burk et al., 2006; Hurst et al., 2010)

‐ Responsive in population with Se status in the lowest range

‐ Plateau in the higher range of Se intakes (> ca. 40–60 μg/day)

AAS: Atomic absorption spectrometer; d: day; ELISA: enzyme‐linked immunosorbent assay; GF: graphite‐furnace; Hb: haemoglobin; HG: Hydride‐generation; ICP‐MS: Inductively coupled plasma mass spectrometry; RBC: red blood cell; Se: selenium.

(a) Serum and plasma selenium concentrations are considered equivalent.

(b) Short term: hours/days; Medium term: weeks; Long term: months.