Place · Level 3
Selenium
硒蛋白的活性中心 · 甲状腺脱碘酶 · 抗氧化系统的微量火花
Story path
Chapter 1
Food · soil matters
Food · soil matters
Selenium is the most typical case of 'soil geochemistry determines nutrition' — the same food can differ by tens of times in selenium content across regions. RDA 55 µg/day, UL 400 µg/day.
Main sources roughly:
Brazil nuts: average 70–90 µg / nut, but with huge variation (20–250 µg / nut even within the same bag, depending on the specific tree's soil) — a rare 'one nut affects your daily Se intake' foodSeafood (tuna, shrimp, sardines, salmon): 30–90 µg / 100 g, small batch variation, a stable sourceMeat + eggs: 10–40 µg / 100 g, depending on feedGrains: heavily dependent on local soil — US Great Plains wheat 50+ µg / 100 g, parts of China < 5 µg
From Heilongjiang to Yunnan in China, there is a low-selenium soil belt. Before the 1970s, Keshan disease (endemic cardiomyopathy) was rampant; after the government fortified table salt with selenium, incidence fell more than 90% — a modern example of a single trace element shaping regional health. Other global low-selenium regions include New Zealand (volcanic soil), parts of northern Europe, sub-Saharan Africa, and the US Northeast and Northwest.
In practice: diversifying food sources across regions is safer than fixing on a single staple; 1–2 Brazil nuts per day cover the RDA at low risk, but eating 10+ per day long-term is not advised (selenosis is real); people who regularly eat seafood, eggs, and meat are almost certainly not deficient.
Main sources roughly:
Brazil nuts: average 70–90 µg / nut, but with huge variation (20–250 µg / nut even within the same bag, depending on the specific tree's soil) — a rare 'one nut affects your daily Se intake' foodSeafood (tuna, shrimp, sardines, salmon): 30–90 µg / 100 g, small batch variation, a stable sourceMeat + eggs: 10–40 µg / 100 g, depending on feedGrains: heavily dependent on local soil — US Great Plains wheat 50+ µg / 100 g, parts of China < 5 µg
From Heilongjiang to Yunnan in China, there is a low-selenium soil belt. Before the 1970s, Keshan disease (endemic cardiomyopathy) was rampant; after the government fortified table salt with selenium, incidence fell more than 90% — a modern example of a single trace element shaping regional health. Other global low-selenium regions include New Zealand (volcanic soil), parts of northern Europe, sub-Saharan Africa, and the US Northeast and Northwest.
In practice: diversifying food sources across regions is safer than fixing on a single staple; 1–2 Brazil nuts per day cover the RDA at low risk, but eating 10+ per day long-term is not advised (selenosis is real); people who regularly eat seafood, eggs, and meat are almost certainly not deficient.
Keshan disease: a geography story
The most classic case of selenium deficiency as a public-health disease comes from China — it's called Keshan disease.Keshan disease was first reported in 1935 in Keshan County, Heilongjiang. Endemic cardiomyopathy — diffuse myocardial degeneration and heart failure — hit children and women of reproductive age hardest. The cause was unknown for decades, until the 1970s when Chinese CDC research confirmed it: local soil was extremely low in selenium (< 0.1 mg/kg), and along the soil → crop → food → human chain, selenium was diluted at every step. After fortifying salt with sodium selenate in high-incidence areas, incidence fell more than 90% within a few years — becoming a modern example of 'a single micronutrient deciding the health of an entire region'.
The Keshan Belt is a low-selenium soil corridor stretching from Northeast China to the Southwest, covering Heilongjiang, Jilin, Liaoning, Inner Mongolia, Hebei, Henan, Shaanxi, Sichuan, Guizhou, Yunnan, and Tibet. The same belt is also high-incidence for Kashin-Beck disease — selenium + iodine double deficiency affecting children's joint cartilage.
Other global low-selenium regions: New Zealand (volcanic rock soil), parts of northern Europe, sub-Saharan Africa, the US Northeast and Northwest.
A larger implication: nutrition isn't something solved just by 'eating more leafy greens' — soil geochemistry is actually upstream. This is also why WHO uses cross-regional dietary diversification as a foundational strategy — a single region's food chain may collectively lack a particular trace element.
Chapter 2
Selenoproteins
Selenoproteins
Selenium is not an ordinary 'plugged-in' mineral — it's written directly into proteins as selenocysteine (Sec, abbreviated U), often called the '21st amino acid'.
This is a curious extension of the genetic code. The classical 20 amino acids are encoded by 64 codons; Sec is encoded by UGA — but UGA is normally a stop codon. Only when the mRNA contains a special SECIS (selenocysteine insertion sequence) structure is UGA 'reinterpreted' as Sec, and this step also requires a special tRNA + EF-Sec + SBP2 working together.
Why did evolution preserve this complex mechanism? Because selenocysteine's selenium reactivity is roughly 1000× that of sulfur (pKa 5.2 vs Cys's 8 — already ionized at physiological pH), making it irreplaceable in fine redox reactions.
The human body has ~25 selenoproteins; the main families include:
GPX (glutathione peroxidases 1–4, 6): clear H₂O₂ and organic peroxidesTXNRD (thioredoxin reductases 1, 2, 3): maintain cellular reduced stateDIO (deiodinases 1, 2, 3): convert T4 to T3 — key to thyroid hormone activation, see next sceneSELENOP: plasma transport and storageSEPP1: neuronal protection
So 'why can micro-changes in trace elements affect the whole body', the answer lies in the selenoprotein family — a small amount of selenium supports an entire family of irreplaceable critical enzymes, which is also why selenium deficiency can clinically involve heart, thyroid, and immunity simultaneously.
This is a curious extension of the genetic code. The classical 20 amino acids are encoded by 64 codons; Sec is encoded by UGA — but UGA is normally a stop codon. Only when the mRNA contains a special SECIS (selenocysteine insertion sequence) structure is UGA 'reinterpreted' as Sec, and this step also requires a special tRNA + EF-Sec + SBP2 working together.
Why did evolution preserve this complex mechanism? Because selenocysteine's selenium reactivity is roughly 1000× that of sulfur (pKa 5.2 vs Cys's 8 — already ionized at physiological pH), making it irreplaceable in fine redox reactions.
The human body has ~25 selenoproteins; the main families include:
GPX (glutathione peroxidases 1–4, 6): clear H₂O₂ and organic peroxidesTXNRD (thioredoxin reductases 1, 2, 3): maintain cellular reduced stateDIO (deiodinases 1, 2, 3): convert T4 to T3 — key to thyroid hormone activation, see next sceneSELENOP: plasma transport and storageSEPP1: neuronal protection
So 'why can micro-changes in trace elements affect the whole body', the answer lies in the selenoprotein family — a small amount of selenium supports an entire family of irreplaceable critical enzymes, which is also why selenium deficiency can clinically involve heart, thyroid, and immunity simultaneously.
The 21st amino acid
The scene body already covered the codon mechanism and the selenoprotein family at the 21st-amino-acid level; here's one experimental-biology footnote.In early Hatfield-lab experiments, researchers tried replacing Sec on key selenoproteins with ordinary Cys, and enzyme activity dropped significantly or even vanished — which is why evolution preserved this expensive SECIS / EF-Sec / SBP2 machinery: critical redox enzymes simply must use Sec, and ordinary Cys can't substitute.
Within the selenoprotein family, GPX clears peroxides, TXNRD maintains cellular reduced state, DIO activates thyroid hormone, SELENOP transports and stores, SEPP1 protects neurons — the remaining dozen or so functions are more distributed. This division of labor explains why, although selenium is needed at only tens of micrograms per day, tiny changes can affect multiple body systems at once.
Chapter 3
Thyroid partner
Thyroid partner
The thyroid is one of the tissues with the highest selenium concentration in the body, because it relies heavily on selenium-containing enzymes.
Two classes of selenoproteins work in parallel in the thyroid. Deiodinases (DIO 1, 2, 3) all contain selenium: DIO 1 + 2 strip an iodine from low-activity T4 to produce high-activity T3, while DIO 3 converts T3 to T2 to inactivate it — together they tune hormone levels. The glutathione peroxidases (GPX) clear by-products: during thyroid hormone synthesis, the thyroid peroxidase: A key enzyme that makes thyroid hormone — in Hashimoto's the immune system often attacks it by mistake. reaction generates H₂O₂, and GPX clears the excess to protect thyroid cells themselves.
So 'iodine + selenium' is a division of labor: iodine is raw material (T4/T3 names already carry 4 / 3 iodines), while selenium is the switch for the hormones plus antioxidant protection for the thyroid. If iodine is short, supplementing selenium blindly doesn't fix the raw-material problem; if selenium is short, the thyroid's capacity to handle oxidative stress drops and it becomes more vulnerable to autoimmune attack.
On Hashimoto's thyroiditis + selenium, there is a moderately solid evidence base. Toulis 2010 meta-analysis (4 RCTs, n=463): 200 µg/day selenomethionine for 3–12 months reduces anti-TPO antibodies by ~21–40%. The 2019 Cochrane review rates the evidence as 'moderate, insufficient to change guidelines'. In clinical guidelines, the European ETA 2013 allows a 6-month trial in low-selenium areas + Hashimoto's patients with very high anti-TPO; the US ATA does not recommend routine use. To be clear, selenium cannot replace levothyroxine — once Hashimoto's progresses to hypothyroidism, hormone replacement is still required.
Practical conclusion: 'selenium saves the thyroid' as a slogan can't be detached from the whole picture — iodine, selenium, autoimmune status, and hormone levels must be read together, and one shouldn't blindly fetishize any single one.
Two classes of selenoproteins work in parallel in the thyroid. Deiodinases (DIO 1, 2, 3) all contain selenium: DIO 1 + 2 strip an iodine from low-activity T4 to produce high-activity T3, while DIO 3 converts T3 to T2 to inactivate it — together they tune hormone levels. The glutathione peroxidases (GPX) clear by-products: during thyroid hormone synthesis, the thyroid peroxidase: A key enzyme that makes thyroid hormone — in Hashimoto's the immune system often attacks it by mistake. reaction generates H₂O₂, and GPX clears the excess to protect thyroid cells themselves.
So 'iodine + selenium' is a division of labor: iodine is raw material (T4/T3 names already carry 4 / 3 iodines), while selenium is the switch for the hormones plus antioxidant protection for the thyroid. If iodine is short, supplementing selenium blindly doesn't fix the raw-material problem; if selenium is short, the thyroid's capacity to handle oxidative stress drops and it becomes more vulnerable to autoimmune attack.
On Hashimoto's thyroiditis + selenium, there is a moderately solid evidence base. Toulis 2010 meta-analysis (4 RCTs, n=463): 200 µg/day selenomethionine for 3–12 months reduces anti-TPO antibodies by ~21–40%. The 2019 Cochrane review rates the evidence as 'moderate, insufficient to change guidelines'. In clinical guidelines, the European ETA 2013 allows a 6-month trial in low-selenium areas + Hashimoto's patients with very high anti-TPO; the US ATA does not recommend routine use. To be clear, selenium cannot replace levothyroxine — once Hashimoto's progresses to hypothyroidism, hormone replacement is still required.
Practical conclusion: 'selenium saves the thyroid' as a slogan can't be detached from the whole picture — iodine, selenium, autoimmune status, and hormone levels must be read together, and one shouldn't blindly fetishize any single one.
Hashimoto's + Se RCT
The scene body already laid out the main Hashimoto's + selenium RCT evidence and guideline stance; here are a few clinical-practice details.Karanikas 2008, a smaller RCT, saw anti-thyroid peroxidase: A key enzyme that makes thyroid hormone — in Hashimoto's the immune system often attacks it by mistake. fall but clinical endpoints like thyroid-stimulating hormone: A pituitary hormone that prods the thyroid to work — it rises when the thyroid is underactive. / T4 improvements were not significant — which is why Cochrane rates the overall evidence as 'moderate' rather than 'definite'. Current Chinese guidelines also do not routinely recommend selenium for Hashimoto's patients.
If you have confirmed Hashimoto's with very high anti-TPO and live in a low-selenium region, you can discuss with your endocrinologist a 6-month trial at 200 µg/day selenomethionine; mind the selenium safety window during use (UL 400 µg/day).
Several edges worth remembering: it cannot replace levothyroxine — once Hashimoto's progresses to hypothyroidism, hormone replacement remains the main line; 'selenium prevents Hashimoto's' currently has no evidence support, but avoiding chronic extreme low selenium is a reasonable target; whether 'selenium + iodine' co-supplementation aggravates immune injury in Hashimoto's patients is still at the theoretical level, with weak evidence — strict physician guidance is enough.
Chapter 4
Immunity & redox
Immunity & redox
Selenium's place in immunity can be summed up in one line: keep oxidative damage limited to where it's supposed to happen.
Immune cells (neutrophils, macrophages) deliberately create oxidative pressure to kill bacteria — H₂O₂, superoxide, hypochlorous acid — but the same weapons damage immune cells themselves and the surrounding normal tissue. GPX glutathione peroxidase and TXNRD thioredoxin reductase reduce the excess peroxides so that damage stays localized around the pathogen; when selenium is insufficient, this regulation blurs and immune dysregulation + chronic inflammation get amplified.
But 'low selenium dulls immunity' does not mean 'high-dose selenium = immunity upgrade'. The most authoritative counter-lesson comes from the SELECT trial (Lippman 2009 JAMA, n=35,533, 7–12 years of follow-up): the design tested whether 200 µg selenium + 400 IU vitamin E in men reduced prostate cancer. The result surprised the field: selenium alone vs placebo, no difference in prostate cancer; vitamin E alone raised prostate cancer +17% (P=0.008); selenium supplementation in men with already adequate baseline selenium raised type 2 diabetes risk by ~7%.
The reasons for failure aren't hard to analyze: these men's baseline selenium was already sufficient (US men average plasma selenium ~130–140 µg/L), and 'adding on top of sufficiency' can be pro-oxidant + interfere with other metabolism; a single nutrient at sustained high doses also doesn't equal a whole dietary pattern.
This history yields two conclusions: first, the assumption that 'plausible mechanism = effective' is unreliable and must be RCT-validated; second, 'correcting deficiency ≠ adding on top of sufficiency' — the core of the U-shaped curve rule. Ordinary men don't need selenium to prevent prostate cancer or to 'boost immunity'; long-term selenium intake shouldn't exceed 200 µg/day.
Immune cells (neutrophils, macrophages) deliberately create oxidative pressure to kill bacteria — H₂O₂, superoxide, hypochlorous acid — but the same weapons damage immune cells themselves and the surrounding normal tissue. GPX glutathione peroxidase and TXNRD thioredoxin reductase reduce the excess peroxides so that damage stays localized around the pathogen; when selenium is insufficient, this regulation blurs and immune dysregulation + chronic inflammation get amplified.
But 'low selenium dulls immunity' does not mean 'high-dose selenium = immunity upgrade'. The most authoritative counter-lesson comes from the SELECT trial (Lippman 2009 JAMA, n=35,533, 7–12 years of follow-up): the design tested whether 200 µg selenium + 400 IU vitamin E in men reduced prostate cancer. The result surprised the field: selenium alone vs placebo, no difference in prostate cancer; vitamin E alone raised prostate cancer +17% (P=0.008); selenium supplementation in men with already adequate baseline selenium raised type 2 diabetes risk by ~7%.
The reasons for failure aren't hard to analyze: these men's baseline selenium was already sufficient (US men average plasma selenium ~130–140 µg/L), and 'adding on top of sufficiency' can be pro-oxidant + interfere with other metabolism; a single nutrient at sustained high doses also doesn't equal a whole dietary pattern.
This history yields two conclusions: first, the assumption that 'plausible mechanism = effective' is unreliable and must be RCT-validated; second, 'correcting deficiency ≠ adding on top of sufficiency' — the core of the U-shaped curve rule. Ordinary men don't need selenium to prevent prostate cancer or to 'boost immunity'; long-term selenium intake shouldn't exceed 200 µg/day.
SELECT trial lesson
The scene body has covered the SELECT trial itself; here's one background note: it was originally designed based on the earlier NPC study (Clark 1996) suggesting selenium may have anti-cancer activity, but with much larger sample size and tighter controls. The result not only failed to see selenium anticancer effects, it also raised prostate cancer +17% in the vitamin-E-alone arm; selenium + vitamin E combined showed no significant difference, with the two partially canceling each other.Large negative results of this kind are real scientific progress — even when commercially 'disappointing', they directly prune away a whole set of repeatedly marketed narratives like 'selenium anti-cancer / vitamin E anti-aging'.
One-sentence takeaway: long-term selenium should stay within the range between RDA and UL (roughly 100–200 µg/day); ordinary men don't need selenium to prevent prostate cancer.
Chapter 5
Narrow window
Narrow window
Selenium's safety window is narrower than most nutrients: RDA 55 µg, UL 400 µg — only about 7×, so long-term overdose really happens.
Typical selenosis (selenium poisoning) presentation:
Brittle, white-streaked, and in severe cases shedding nailsDry hair, hair lossMetallic or garlic taste in mouth or breath (dimethylselenide exhaled via the lungs is a relatively specific signal)Nausea, diarrheaPeripheral neuropathy with numbness and tinglingLiver dysfunction in extreme cases
Brazil nuts are 'the most drug-like food' among nuts: average 70–90 µg per nut, with extreme values from 20 µg to 250 µg. 1–2 per day will likely cover the RDA at low risk; 3–5 per day may approach or exceed the UL; eating 10+ per day chronically has real reported cases of selenosis.
Combined with the SELECT trial follow-up from the previous scene — selenium supplements raised T2D risk by ~7% in those with already adequate baseline selenium — 'a bit more never hurts' is especially wrong for selenium.
A few practical points:
People eating seafood, eggs, meat, and grains normally mostly don't need extra seleniumIf supplementing, prefer the stable form: food-source selenomethionine over inorganic sodium seleniteGenuine indications for selenium supplementation are limited: Hashimoto's patient + selenium-deficient region + very high anti-thyroid peroxidase: A key enzyme that makes thyroid hormone — in Hashimoto's the immune system often attacks it by mistake., after discussion with an endocrinologistDon't treat Brazil nuts as a daily snack eaten by the handful — among nuts it sits in a unique dose-risk tier
Typical selenosis (selenium poisoning) presentation:
Brittle, white-streaked, and in severe cases shedding nailsDry hair, hair lossMetallic or garlic taste in mouth or breath (dimethylselenide exhaled via the lungs is a relatively specific signal)Nausea, diarrheaPeripheral neuropathy with numbness and tinglingLiver dysfunction in extreme cases
Brazil nuts are 'the most drug-like food' among nuts: average 70–90 µg per nut, with extreme values from 20 µg to 250 µg. 1–2 per day will likely cover the RDA at low risk; 3–5 per day may approach or exceed the UL; eating 10+ per day chronically has real reported cases of selenosis.
Combined with the SELECT trial follow-up from the previous scene — selenium supplements raised T2D risk by ~7% in those with already adequate baseline selenium — 'a bit more never hurts' is especially wrong for selenium.
A few practical points:
People eating seafood, eggs, meat, and grains normally mostly don't need extra seleniumIf supplementing, prefer the stable form: food-source selenomethionine over inorganic sodium seleniteGenuine indications for selenium supplementation are limited: Hashimoto's patient + selenium-deficient region + very high anti-thyroid peroxidase: A key enzyme that makes thyroid hormone — in Hashimoto's the immune system often attacks it by mistake., after discussion with an endocrinologistDon't treat Brazil nuts as a daily snack eaten by the handful — among nuts it sits in a unique dose-risk tier
Brazil nuts: food-as-drug
Brazil nuts (*Bertholletia excelsa*) are the highest natural-selenium food and one of the few where 'a single nut's dose can swing a day's selenium intake'.Content averages 70–90 µg / nut, but variation is huge: nuts in the same bag can range from 20 µg to 250 µg depending on the tree's soil; a 100 g bag (~50 nuts) can hold total selenium between 1,500–4,500 µg. Against the RDA 55 µg/day and UL 400 µg/day, 1–2 per day will likely cover the RDA at low risk; 3–5 may approach or exceed the UL; 10+ per day chronically has real reported selenosis cases.
Selenosis symptoms include brittle nails with white streaks or even shedding, dry hair and alopecia, metallic or garlic taste in mouth or breath (dimethylselenide exhaled via the lungs), nausea and diarrhea, neuropathic numbness and tingling, and liver dysfunction in extreme cases.
A few practical suggestions: don't snack on Brazil nuts by the handful daily — among nuts they're in their own dose-risk tier; 2 nuts/day is often recommended as a 'selenium-supplement substitute', but watch for batch variation; seafood (tuna, shrimp, sardines), meat, eggs, and grains are more stable selenium sources with smaller batch variation; if you really need to supplement, selenomethionine has better absorption and bioavailability than sodium selenite, but daily-life populations usually don't need supplementation at all.
In special scenarios, some Hashimoto's / Graves' patients see thyroid peroxidase: A key enzyme that makes thyroid hormone — in Hashimoto's the immune system often attacks it by mistake. antibody reduction with 200 µg/day sodium selenate (Toulis 2010 meta-analysis, moderate clinical evidence); this is a limited use in diagnosed patients, not 'ordinary people taking selenium to prevent thyroid disease'.