Objective: To assess the absorption of dietary selenium in humans, especially of milk selenium.
Design: 1-day meal studies in subjects with ileostomy.
Setting: Hospital outpatient clinics.
Subjects: Three subjects in the pilot study and nine subjects in the main study (eight men/ four women).
Interventions: Different beverages, 1 l/day, were given in addition to basal diets (soft drink, 1 week; low-fat milk, 3 weeks; fermented low-fat milk, 3 weeks and soft drink, 1 week). Ileostomy effluents were collected during the last 2 days in each of the four periods.
Results: On days when the subjects were given 1 l of low-fat milk, the estimated fractional absorption of total dietary selenium was 65.5 (2.3)% (mean (s.d.), n=18), which was similar to the value when fermented low-fat milk was given (64.1 (3.2)%). However, both the calculated amount of milk selenium absorbed (10.9 (2.4) vs 9.4 (1.7) μg selenium) and its fractional absorption (73.3 (16.1) vs 64.1 (11.2)%, n=18) were significantly higher for milk than for fermented milk.
Conclusions: Selenium from milk and other sources is well absorbed in subjects with ileostomy. The real absorption may be even higher than the values shown.
Sponsorship: The Swedish Dairy Association, the Swedish Farmers' Foundation for Agricultural Research, the Crafoord Foundation and Lund University Hospital.
The role of selenium for the maintenance of health and the prevention of several diseases is probably more important than realised previously. Besides, its putative role in cancer prevention (Clark et al, 1996; Ip, 1998), selenium is now linked to a number of physiological functions as reviewed by Rayman (2000). Moreover, several new selenoproteins with unknown function have been detected recently (Kryukov & Gladyshev, 2002; Lescure et al, 2002).
The bioavailability of selenium from several foods has been investigated previously mainly in animal models but also in a few human studies (Levander et al, 1983; Mutanen, 1986; Fairweather-Tait, 1997). The investigations were carried out using different diets and food items (Van der Torre et al, 1991; Meltzer et al, 1993; Shi & Spallholz, 1994), and several kinds of selenium supplements (Levander et al, 1983; Luo et al, 1985; Alfthan et al, 2000). No study in humans has been performed on the bioavailability of selenium from milk.
As a major food in the Swedish diet, milk products provide an average of 17% of the total selenium intake for the population in Sweden (Becker, 2000). The bioavailability of selenium from milk was studied previously in rats using the plasma selenium concentration and glutathione peroxidase (GSHPx) activity as biomarkers (Mutanen et al, 1986), showing that milk selenium was almost as available as the reference compound selenite. In contrast, in vitro simulated gastrointestinal digestion methods (Shen et al, 1996) indicated its bioavailability from different milks to be 2–11%. Since these studies did not give consistent results, the present study was designed to estimate the gastrointestinal absorption of selenium from milk products in humans using another approach, the ileostomy model.
Materials and methods
Subjects and experimental design
Three subjects with ileostomy (subject 1=F, 44; 2=M, 22; 3= F, 44) were included in a 1-day pilot experiment (Hertervig, 2000). They had undergone subtotal colectomy with the creation of an ileostomy because of refractory inflammatory bowel disease, at least 6 months prior to the investigation. On the day of the experiment, the three subjects were given a low selenium basal diet during 8 h starting with breakfast at 08.00 (Table 1). At breakfast, two of them were also given 200 ml of milk. The ileostomy effluents of each subject were collected at time 0 and then at 2-h intervals for the 8 h. The ileostomy effluents and duplicates of the basal diet were freeze-dried and weighed, and were analysed for selenium content.
Nine subjects with ileostomy (average age=56.3 y, two females and seven males) in healthy condition were included in this experiment, which was described previously (Tidehag et al, 1995). The subjects had been proctocolectomised due to ulcerative colitis 3–20 y before this study. The experiment lasted for 8 weeks, and it was divided into four periods of 1, 3, 3 and 1 weeks, respectively. The subjects were asked to consume 1000 ml of beverage every day in addition to their basal diets and coffee. The beverage volume was evenly consumed in four servings together with the three main meals and an evening snack. In periods 1 and 4, the beverage was a colourless soft drink and in periods 2 and 3, that is, the 3-week test periods, the beverages consisted of low-fat milk and fermented low-fat milk, respectively, which were given in a randomised design. Most of the time the subjects lived at home and had their usual diets with some restrictions, but during the last 2 days in each period, they were given controlled test diets at the research ward (Table 2a and b). For one basal portion per day, the major carbohydrate sources were wheat crisp bread (79 g), potato (160 g) and sugar (17.5 g), the major protein sources were ham (60 g), minced beef (150 g), fillet of pork (150 g) and egg (25 g), while the major fat source was margarine (51 g). Three subjects consumed 1.5 basal portions on each test day, three consumed 1.25 portions, two consumed 1.0 portion and one 0.75 portions based on their respective estimated energy requirements. For each subject, the ileostomy effluents were collected during the two test days in each period, pooled for each 24-h interval, freeze-dried and weighed. The basal diets, test beverages (1000 ml each) and coffee (10 g per day) were also freeze-dried and weighed. All samples were assayed for selenium content.
The selenium content of the samples were analysed using hydride generation graphite furnace atomic absorption spectrometry (HG-GF-AAS, Perkin-Elmer Aanalyst 800) in combination with flow-injection analysis (FIAS-400), as described by Daun et al (2001). As controls, the certified reference materials lyophilised bovine liver (NIST 1577b) or bovine muscle (CRM 184) were analysed in every assay. The within-day and between-day imprecision in the analysis of one basal diet sample was 2.0 and 6.2% (CV), respectively.
The studies were approved by the Research Ethics Committees at the University of Umeå and at Lund University.
Calculations and statistical methods
The total selenium intake was calculated from the selenium content of the basal diet portion and the respective beverages. The apparent selenium absorption was calculated as intake minus ileal excretion. The fractional absorption of selenium was calculated as 100 × (intake−excretion)/intake. It was realised that the selenium absorption might be underestimated in this manner since endogenous selenium compounds which are secreted into the gastrointestinal lumen but not absorbed in the small intestine might increase selenium excretion, resulting in a lower estimate of the absorption of exogenous selenium. The absorption of milk selenium, expressed as μg/day, was calculated as total selenium absorption during milk periods minus total selenium absorption during non-milk periods (mean of 4 days). The fractional absorption of milk selenium was then calculated as 100 × absorption of selenium from milk/intake of milk selenium. The statistical significances of the differences between absorption data from milk, fermented milk and control periods were calculated using analysis of variance. Linear correlation coefficients were computed.
This pilot study was made to assess the feasibility of the ileostomy model for the measurement of selenium absorption. The selenium content of the basal diet on the test day was 9.1 μg and the additional 200 ml of milk contained 3.6 μg selenium contributing approx. 28% of the total selenium intake. The subject consuming a non-milk diet showed a lower level of both total selenium absorption and fractional absorption (62%) compared to the other two subjects (74–82%) who took 200 ml milk with breakfast (Table 3). These data suggested that both selenium from the mainly vegetarian meals and milk selenium were well absorbed. It was also concluded that for a better assessment of the absorption of selenium from milk, a study design with a larger intake of milk would be needed.
The selenium content of the basal test diet (1.0. portion) was 52 μg (Table 2b). The diet was relatively rich in selenium due to its high content of meat and egg. The soft drink used as control beverage and the coffee provided only <0.5 μg of selenium per day. Low-fat milk and fermented low-fat milk contained approx. 15 μg selenium per litre contributing 16–27% of the total selenium intake depending on the size of the diet portions (1.5–0.75) consumed by the subjects. When 1000 ml of milk or fermented milk were consumed in addition to the basal test diet, the daily dietary selenium intake ranged from 54 to 93 μg.
The difference between total selenium intake and the selenium content of the pooled 24-h ileostomy effluents was used as an indicator of the absorption of selenium in the gastrointestinal tract. It ranged between 25.9 and 55.5 μg during the non-milk days and it was 35.7–61.9 and 36.6–60.6 μg during the milk and fermented milk periods, respectively. The fractional absorption expressed as percent of total selenium intake was also calculated for each test day (Table 4). The mean values for different days ranged between 60.7 and 66.2%, but there were no significant differences among the four test periods.
The difference between the amounts of selenium absorbed in milk periods and in soft drink periods was used as an indicator of the absorption of milk selenium (Figure 1). The average amount of selenium absorbed from low-fat milk was calculated as 10.9 (2.4) μg/day (mean (s.d.) n=18), which was significantly higher (P=0.007) than that absorbed from fermented low-fat milk, 9.4 (1.7) μg/day. The fractional absorption of selenium from low-fat milk was significantly higher (P=0.010) than the value for fermented low-fat milk, 73.3 (16.1) and 64.1 (11.2)%, respectively.
The fractional absorption of total selenium was found to be inversely associated to the basal portion size in all the four periods, the correlation coefficients being −0.54 (P=0.022), −0.43 (P=0.072), −0.70 (P<0.001) and −0.47 (P=0.049) for the first control, milk, fermented milk and second control periods, respectively.
Several techniques have been used to assess the absorption of trace elements. Previously, the intestinal balance alternative of the chemical balance technique was used in studies of the absorption of iron and calcium (Tidehag et al, 1995). In this manner, the problems with incomplete collection of faeces and with assessment of the demarcation of collection periods in relation to intake are minimised. Since only few studies on selenium absorption have been performed in humans, it was interesting to use this method also for studies of selenium absorption. Previous balance studies showed that the apparent absorption of selenium estimated as intake minus faecal output was in the range 37–72% (Robinson & Thomson, 1983).
Since the ileostomy model has not previously been used for assessment of selenium absorption, a pilot study using low-selenium meals together with a moderate amount of milk was performed. The results indicated a relatively high absorption of selenium both from milk and the total diet. It was also clear that for a more detailed assessment of the absorption of selenium from milk, a study design with a larger intake of milk would be needed.
The estimated absorption of total selenium was in the range 61–66% for different days of study and it was 73 and 64% for selenium from low-fat milk and fermented milk, respectively. This is comparable to results from an isotopic tracer (74Se) study, in which selenite absorption from a bovine milk-based formula in children was 64.2 (14.6)% (mean (s.d.), Solomons et al, 1986). In another study using an extrinsic stable isotope the absorption of selenium from a standard formula was 91% and from a selenium supplemented formula 86%, in very-low-birth-weight infants (Ehrenkranz et al, 1991). The net absorption of selenium using balance data was approx 70%. Very recently, Van Dael et al (2002) found that 97% of selenate and 73% of selenite added to a milk-based infant formula was absorbed in infants. Very few studies of selenium absorption have been performed using intrinsically labelled foods. With the aid of this technique the absorption of selenium from poultry meat, egg white and egg yolk was found to be 71, 54 and 78%, respectively (Christensen et al, 1983; Sirichakwal et al, 1985). No study on selenium absorption using intrinsically labelled milk seems to have been performed. Although the cited data on selenium absorption from milk and milk-based products may not be strictly comparable due to the use of different tracers, they all fall within the range 61–97%, the data for chicken meat and egg yolk also falling within this range.
In animal studies the extent of selenium absorption has instead been expressed in relation to that of a reference substance, usually selenite. In rats the bioavailability of milk selenium originating from selenium-supplemented cows and normal American milk was only slightly lower than that of selenite using plasma selenium concentration and plasma/liver glutathione peroxidase activity as biomarkers (Mutanen et al, 1986). The bioavailability of selenium from different milks was however much lower, 2–11%, as assessed in an in vitro simulated gastrointestinal digestion study (Shen et al, 1996). Further studies from that research group showed the selenium availability from bovine milk to be largely influenced by protein digestibility. The up to 10-fold difference in selenium bioavailability obtained in the ileostomy design and the simulation model is probably explained by a more complete digestion in the real gastrointestinal tract. Moreover, other mechanisms of absorption operate under physiological conditions than in the simulated system, where only in vitro digestion and diffusion was measured. In addition, since selenomethionine and selenate showed higher diffusion rates than selenocysteine and selenite in vitro (Shen et al, 1997), the results from both experimental designs may be influenced by the occurrence of different forms of selenium in food.
Experiment 2 showed a somewhat higher selenium absorption from low-fat milk than from fermented low-fat milk. The fermentation may have altered the composition of selenocompounds in milk, thus affecting their digestibility and absorbability, but further studies are necessary to document the effects of fermentation on selenium bioavailability. During manufacture of fermented milk usually heat treatment at a higher temperature (>90°C) is used than for unfermented milk, which will cause denaturation of glutathione peroxidase in milk (Lindmark Månsson et al, 2001). Also, other heat-induced changes in selenium-containing compounds may explain the somewhat lower bioavailability of selenium in fermented milk. Regarding other factors associated with selenium absorption, it is unclear why there was an inverse relationship between its fractional absorption and the diet portion size. It may be related to the occurrence of a dietary component(s) in the standard diet inhibiting selenium absorption, but such an effect would be expected to be counteracted by the fact that also the selenium intake increased with increasing diet portion size. As this relationship seemed to be more evident in the fermented-low-fat-milk test period than in other periods, the possible interactions between selenium and other compounds may be more important in the fermented milk. It should also be noted that this inverse relationship was observed for total selenium only but not for the calculated absorption of milk selenium.
In the two experiments of this study the amount of selenium absorbed varied at least two-fold depending mainly on the dietary intake of selenium, although the fractional absorption of selenium decreased somewhat with increasing diet portion size. This supports previous findings that the gut has little homeostatical control over selenium absorption, and instead the renal excretion of selenium may be a major means of controlling whole-body selenium content (King, 2001). The mean values for absorption of total selenium and milk selenium on different days varied between 61 and 77% indicating a high bioavailability in this first study of selenium absorption using the ileostomy model. The true absorption may be even higher since the possible occurrence of endogenously secreted but not absorbed selenium will decrease the calculated value of fractional absorption. To estimate this contribution a refined experimental design of the ileostomy model using milk intrinsically labelled with stable isotopes would be necessary (Fairweather-Tait, 1997). Moreover, by definition, studies in subjects with ileostomy will not include any possible absorption or secretion of selenium in the large bowel.
Alfthan G, Xu GL, Tan WH, Aro A, Wu J, Yang YX, Liang WS, Xue WL & Kong LH (2000): Selenium supplementation of children in a selenium-deficient area in China: blood selenium levels and glutathione peroxidase activities. Biol. Trace Elem. Res. 73, 113–125.
Becker W (2000): Vilka är källorna till våra näringsämnen? Vår Föda 3, 16–20.
Clark LC, Combs GF, Turnbull BW, Slate EH, Chalker DK, Chow J, Davis LS, Glover RA, Graham GF, Gross EG, Krongrad A, Lesher JL, Park K, Sanders BB, Smith CL & Taylor R (1996): Effects of selenium supplementation for cancer prevention in patients with carcinoma of the skin. JAMA 276, 1957–1963.
Christensen MJ, Janghorbani M, Steinke FH, Istfan N & Young VR (1983): Simultaneous determination of absorption of selenium from poultry meat and selenite in young men: application of a triple stable-isotope method. Br. J. Nutr. 50, 43–50.
Daun C, Johansson M, Önning G & Åkesson B (2001): Glutathione peroxidase activity, tissue and soluble selenium content in beef and pork in relation to meat ageing and pig RN phenotype. Food Chem. 73, 313–319.
Ehrenkranz RA, Gettner PA, Nelli CM, Sherwonit EA, Williams JE, Ting BT & Janghorbani M (1991): Selenium absorption and retention by very-low-birth-weight infants studied with the extrinsic stable isotope tag 74Se. J. Pediatr. Gastroenterol. Nutr. 13, 125–133.
Fairweather-Tait SJ (1997): Bioavailability of selenium. Eur. J. Clin. Nutr. 51, S20–S23.
Hertervig E (2000): Alkaline sphingomyelinase a potential inhibitor in colorectal carcinogenesis. Ph.D thesis, Lund University, Sweden.
Ip C (1998): Lessons from basic research in selenium and cancer prevention. J. Nutr. 128, 1845–1854.
King JC (2001): Effect of reproduction on the bioavailability of calcium, zinc and selenium. J. Nutr. 131, 1355S–1357S.
Kryukov GV & Gladyshev VN (2002): Mammalian selenoprotein gene signature: identification and functional analysis of selenoprotein genes using bioinformatics methods. Methods Enzymol 347, 84–100.
Lescure A Gautheret D & Krol A (2002): Novel selenoproteins identified from genomic sequence data. Methods Enzymol 347, 57–70.
Levander OA, Alfthan G, Arvilommi H, Gref CG, Huttunen JK, Kataja M, Koivistoinen P & Pikkarainen J (1983): Bioavailability of selenium of Finnish men as assessed by platelet glutathione peroxidase activity and other blood parameters. Am. J. Clin. Nutr. 37, 887–897.
Lindmark-Månsson H, Chen J, Paulsson M, Aldén G, Ren B, Ladenstein R & Åkesson B (2001): The effect of storage and heat treatment on glutathione peroxidase in bovine milk and whey. Int. Dairy J. 11, 71–81.
Luo XM, Wei HJ Yang CL, Xing J, Liu X, Qiao CH, Feng YM, Liu J, Liu YX, Wu Q, Liu X, Guo JS, Stoecker BJ, Spallholz JE & Yang SP (1985): Bioavailability of selenium to residents in a low-selenium area of China. Am. J. Clin. Nutr. 42, 439–448.
Meltzer HM, Bibow K, Paulsen IT, Mundal HH, Norheim G & Holm H . (1993): Different bioavailability in humans of wheat and fish selenium as measured by blood platelet response to increased dietary Se. Biol. Trace Elem. Res. 36, 229–241.
Mutanen M (1986): Bioavailability of selenium. Ann. Clin. Res. 18, 48–54.
Mutanen M, Aspila P & Mykkänen HM (1986): Bioavailability to rats of selenium in milk of cows fed sodium selenite or selenited barley. Ann. Nutr. Metab. 30, 183–188
Rayman MP (2000): The importance of selenium to human health. Lancet 356, 233–241.
Robinson MF & Thomson CD (1983): The role of selenium in the diet. Nutr. Abstr. Rev. 53, 3–26
Shen L, Van Dael P, Luten J & Deelstra H (1996): Estimation of selenium bioavailability from human, cows', goat and sheep milk by an in vitro method. Int. J. Food Sci. Nutr. 47, 75–81.
Shen L, Van Dyck K, Luten J & Deelstra H (1997): Diffusibility of selenate, selenite, seleno-methionine, and seleno-cystine during simulated gastrointestinal digestion. Biol. Trace Elem. Res. 58, 55–63.
Shi B & Spallholz JE (1994): Selenium from beef is highly bioavailable as assessed by liver glutathione peroxidase (EC 1.11.19) activity and tissue selenium. Br. J. Nutr. 72, 873–881.
Sirichakwal PP, Young VR & Janghorbani M (1985): Absorption and retention of selenium from intrinsically labelled egg and selenite as determined by stable isotope studies in humans. Am. J. Clin. Nutr. 41, 264–269.
Solomons NW, Torun B, Janghorbani M, Christensen MJ, Young VR & Steinke FH (1986): Absorption of selenium from milk protein and isolated soy protein formulas in preschool children: studies using stable isotope tracer 74Se. J. Pediatr. Gastroenterol. Nutr. 5, 122–126.
Tidehag P, Sandberg AS, Hallmans G, Wing K, Turk M, Holm S & Grahn E (1995): Effect of milk and fermented milk on iron absorption in ileostomy subjects. Am. J. Clin. Nutr. 62, 1234–1238.
Van Dael P, Davidsson L, Ziegler EE, Fay LB & Barclay D (2002): Comparison of selenite and selenate apparent absorption and retention in infants using stable isotope methodology. Pediatr. Res. 51, 71–75
Van der Torre HW, Van Dokkum W, Schaafsma G, Wedel M & Ockhuizen T (1991): Effect of various levels of selenium in wheat and meat on blood Se status indices and on Se balance in Dutch men. Br. J. Nutr. 65, 69–80.
We thank the study subjects for their willingness to participate. The study was supported by grants from the Swedish Dairy Association, the Swedish Farmers' Foundation for Agricultural Research, the Crafoord Foundation and Lund University Hospital.
About this article
In vivo and in vitro testing for selenium and selenium compounds bioavailability assessment in foodstuff
Critical Reviews in Food Science and Nutrition (2017)
Dietary habits and selenium, glutathione peroxidase and total antioxidant status in the serum of patients with relapsing-remitting multiple sclerosis
Nutrition Journal (2014)
The American Journal of Clinical Nutrition (2010)
A short-term intervention trial with selenate, selenium-enriched yeast and selenium-enriched milk: effects on oxidative defence regulation
British Journal of Nutrition (2008)