Following a request from the European Commission, the EFSA Panel on Dietetic Products, Nutrition and Allergies (NDA) was asked to deliver a scientific opinion on Dietary Reference Values (DRVs) for the European population, including selenium.
In the diet, selenium is mainly present in organic compounds, as l-selenomethionine and l-selenocysteine, with lower amounts in inorganic compounds, as selenate and selenite. Because quantification and speciation of selenium in foods is complex and because there is considerable variation in the selenium content of foods, food composition tables are often inaccurate, resulting in imprecise estimates of selenium intake.
A total of 25 selenoproteins with a variety of functions, including antioxidant effects, T-cell immunity, thyroid hormone metabolism, selenium homeostasis and transport, and skeletal and cardiac muscle metabolism, have been identified in humans. Selenoprotein P (SEPP1) plays a central role in selenium supply to tissues and participates in the regulation of selenium metabolism in the organism.
Selenium in its various forms appears to be well absorbed from the diet. Upon absorption, selenocysteine, selenate and selenite are available for the synthesis of selenoproteins. Selenomethionine is non-specifically integrated into the methionine pool and can substitute for methionine in proteins. Selenomethionine may also be converted to selenocysteine and enter the functional selenium body pool. The production of methylated selenium compounds in the liver, which are excreted predominantly in the urine, participates in the regulation of selenium metabolism in the organism.
Selenium deficiency affects the expression and function of selenoproteins and has been involved in the degeneration of organs and tissues leading to the manifestation of Keshan and Kashin-Beck diseases.
Plasma selenium includes selenium in selenoproteins (the functional pool of selenium), and other plasma proteins in which selenomethionine non-specifically substitutes for methionine. Thus, plasma selenium is not a direct marker of the functional selenium body pool. Measures of glutathione peroxidases (GPxs) activity can be used as a biomarker of selenium function. However, the activity of GPxs reaches a steady state with levels of selenium intake that are lower than those required for the levelling off of SEPP1. The latter is considered the most informative biomarker of selenium function on the basis of its role in selenium transport and metabolism and its response to different forms of selenium intake. Intervention studies using different levels of selenium intake showed that plasma SEPP1 concentration levels off in response to increasing doses of selenium. The levelling off of plasma SEPP1 was considered to be indicative of an adequate supply of selenium to all tissues and to reflect saturation of the functional selenium body pool, ensuring that selenium requirement is met. This criterion was used for establishing DRVs for selenium in adults.
Evidence from human studies on the relationship between selenium intake and plasma SEPP1 concentration was reviewed. The Panel noted uncertainties with respect to estimates of background selenium intake in most studies. Habitual selenium intakes of 50–60 µg/day were not sufficient for SEPP1 concentration to reach a plateau in Finnish individuals, while selenium intakes of 100 µg/day and above were consistently associated with plasma SEPP1 concentration at a plateau in population groups from Finland, the UK and the USA. In a study in healthy individuals from New Zealand, selenium intakes of around 60–70 µg/day were required for SEPP1 concentration to level off. Although this was the only study that quantified background selenium intake from the analysed selenium content of consumed foods, the Panel noted the large variability in the results of this study. In another study among Chinese subjects, a selenium intake of 0.85 µg/kg body weight per day led to the levelling off of plasma SEPP1 concentration. The Panel noted, however, that there were uncertainties related to the intake estimates and to the extrapolation of results from Chinese individuals to the European population.The Panel also noted uncertainties in extrapolating values derived from studies that administered selenium as l-selenomethionine to dietary selenium including other forms of selenium.
Given the uncertainties in available data on the relationship between total selenium intake and SEPP1 concentration, they were considered insufficient to derive an Average Requirement for selenium in adults. Instead, an Adequate Intake (AI) of 70 µg/day for adult men and women was set. A review of observational studies and randomised controlled trials that investigated the relationship between selenium and health outcomes did not provide evidence for additional benefits associated with selenium intake beyond that required for the levelling off of SEPP1.
No specific indicators of selenium requirements were available for infants, children or adolescents.
For infants aged 7–11 months, an AI of 15 µg/day was derived by extrapolating upwards from the estimated selenium intake with breast milk of younger exclusively breast-fed infants and taking into account differences in reference body weights. For children and adolescents, the AIs for selenium were extrapolated from the AI for adults by isometric scaling and application of a growth factor. The AIs range from 15 µg/day for children aged one to three years to 70 µg/day for adolescents aged 15–17 years.
There is evidence suggesting adaptive changes in the metabolism of selenium during pregnancy, and it was considered that these changes cover the additional selenium needs during this period. The Panel proposes that the AI set for adult women also applies to pregnancy. Based on an average amount of selenium secreted in breast milk of 12 μg/day and an absorption efficiency of 70 % from usual diets, an additional selenium intake of 15 µg/day was considered to replace these losses. Thus, an AI of 85 μg/day is proposed for lactating women.