Se intakes vary country by country, ranging from a mean of 40 µg/day in Europe to 134 µg/day in the US [12], leading to different dietary reference intakes for adults, from 55 µg/d in US [44] to 70 µg/d in Europe [45, 46]. A study in Taiwan that investigated the amount of Se, vitamin E, and other nutrients in balanced diets planned according to the national RDA reported Se content among six diet samples to be 112 ± 7 µg/d, well above the recommended 55 µg/d for Taiwanese adults [47]. This means that adherence to dietary guidelines can prevent the general Taiwanese population from Se deficiency. However, to date, no large-scale study has been conducted to investigate Se intakes in the Taiwanese population. Only one study by Han (2022), which estimated Se intakes through food frequency questionnaires in 21 Taiwanese adults (aged 18–31 y), reported the mean intake to be 44.5 µg/day (ranged 1.1–324.9 µg/day) [48]. Discrepancies also exist in ranges considered “normal” by different hospitals and clinical laboratories in Taiwan. It is therefore necessary to establish a consensus reference value for Se level in Taiwan to enable the assessment of its relationship to disease prevention and health promotion in the context of the local population.
We calculated the combined mean and SD (Table 2) from included sources of data and estimated the range of Se levels in the Taiwanese adult population. The mean serum Se level was 182.8 ± 124.3 µg/L, which was similar to whole blood Se of US adults from the National Health and Nutrition Examination Survey (NHANES) 2011–2014 [49]. However, the value was relatively higher than a previous Taiwan-based study (110.9 ± 21.5 µg/L) [20]. This was probably contributed by the employment of a different bioanalytical method, different subjects’ characteristics including a different age range (> 15 years old in the previous study compared to ≥ 19 years old in this study), county of residence where regional variation of soil Se content plays a role [50], and increased consumption of whole grains, nuts, and fish over the years since the previous study [51]. Studies on the endemic selenium intoxication in humans estimated an average intake of 5,000 µg/day and blood Se level of 3,200 µg/L in affected individuals in Hubei Province, China [52]. The lowest-observed-adverse-effect-level (LOAEL) was reported to be 910 µg/day (corresponding blood Se level was 1,054 µg/L) and the none-observed-adverse-effect-level (NOAEL) was 819 µg/day [53]. For a maximum daily safe intake of 400 µg/day, the corresponding levels of Se in whole blood and plasma were 559 µg/L and 327 µg/L, respectively [54]. From these literature, a serum/plasma Se level 182.8 µg/L, much lower than the toxic concentration, is regarded to be safe.
Based on the quartile analysis (Table 3), we categorized Se status into low/suboptimal, normal/adequate, and high corresponding to Se levels in the first, second to third, and fourth quartile, respectively. This approach defined low Se levels to be < 98.8 µg/L, suggesting the population having Se levels lower than approximately 100 µg/L is likely prone to compromised health due to Se deficiency-related issues, especially those experiencing critical conditions such as severe trauma and septic shock because of increased oxidative stress and acute phase response that deplete Se stores in the body [55]. In addition, despite large variations in the reported Se values across the studies (Online Resource Tables S1 and S2), lower Se levels were consistently observed in the disease groups compared to their respective control groups, and lower Se levels were associated with increased risk of disease. Moreover, many of these studies were carried out in the elderly population, which hints that this group is particularly prone to Se deficiency and its effects, such as compromised immunity and impaired physical function.
Low levels of Se and Gpx3 protein and bioactivity have been observed in sepsis patients and predict intensive care unit (ICU) and hospital mortality in patients with septic shock [56–58]. Se levels below the normal range due to long-term total parenteral nutrition has long been identified in clinical nutrition before the 1990s [59], leading to symptoms including muscle weakness [60], whitened nail beds [61], and neurological deficit [62], among others. Since then, Se has been recognized as an essential component of parenteral solutions [63, 64] to prevent these symptoms. Se supplementation in the critically ill has been reported to decrease mortality and the lengths of stay in ICU and hospital, possibly attributed to increased Se levels which were positively correlated with serum GPx3 activity and negatively correlated with inflammatory cytokines interleukin-1 beta (IL-1β) and IL-6 [65–68]. Increased serum Se levels were also associated with reduced risk of liver fibrosis and all-cause mortality [69], as well as improved cardiovascular health and decreased cardiovascular mortality when administered together with antioxidants [17, 70]. Therefore, achieving higher levels of Se may offer therapeutic dosages in a variety of illnesses and may be critical for survival in the critically ill with low baseline Se [71]. A study that evaluated Se status and all-cause and cancer mortality in US adults has observed serum Se of 130–150 µg/L to be associated with minimal mortality [72], whereas an ideal Se level around 125 µg/L has been suggested for disease risk reduction [13]. These values are in line with the “adequate” Se levels within the second quartile (98.8–182.8 µg/L) of our calculation.
Nonetheless, a Taiwan-based study concluded higher serum Se level (> 104.5 µg/L) is associated with a higher diabetes prevalence, although causal relationship cannot be established, and the association may be affected by potentially unidentified confounders [30]. Other adverse effects related to higher Se levels in the third and fourth quartiles of the calculated range have not been reported in Taiwan, and this could be a topic for future endeavor. In short, this new range of Se levels can be used to suggest Se status and may be applied by physicians in Taiwan specifically for the identification of individuals who may need additional Se via dietary intake, supplementation, or parenteral nutrition and to inform clinical practice based on the evaluation.
A few limitations are present in this study. Due to the large SD of the NAHSIT data, the range can only be calculated by the combined mean ± 1SD, which, according to the empirical rule of 68-95-99.7%, encompasses only 68% of the population. However, NAHSIT was included because it is a government-initiated nationwide study using a stratified, multistage cluster design and probability proportional to size sampling to provide data representing the noninstitutionalized Taiwanese population, in addition to considerations for geographical and season variations. Due to this restraint, the results obtained in this study may be applicable to a smaller proportion of the population. If possible, a population-based study on interactions of genetic variation and oxidative stress pathways associated with Se may provide useful insights to delineate the relationship between Se status and health in the Taiwanese population [73]. In addition, serum/plasma Se levels indicate the systemic content of the mineral, whereas the concentration and activity of blood or tissue selenoproteins, specifically GPx and SEPP1, reveal the amount of Se that is available to maintain physiological wellbeing [74]. This study did not assess concentration and activity of specific selenoproteins in the Taiwanese population, but inclusion of these biomarkers can complement the categorization of Se status.