Silicon is a chemical element found in food and water. Foods hold it as silica and silicates; water holds orthosilicic acid, EFSA notes. Whether the human body needs it has not been proven. Supplements use choline-stabilized orthosilicic acid (ch-OSA), monomethylsilanetriol (MMST), silicon dioxide (silica), or horsetail and bamboo extracts.
Terms and sources
- silica and silicates
- EFSA, 2004: "Silicon occurs naturally in foods as silicon dioxide (silica) and silicates".
- orthosilicic acid
- Written Si(OH)4. EFSA, 2004: "Silicon in water is present as ortho-silicic acid Si(OH)4".
- has not been proven
- National Academies, 2001: "A biological role of silicon in humans has not yet been identified". EFSA, 2004: "Silicon has not been shown to be essential for humans".
- Supplements use
- Forms used in the trials below, in EU lists and in Italy's list of plants for supplements. The forms section below compares them.
Silicon in the body
In blood, silicon exists almost entirely as silicic acid, the National Academies note. As much as 50 percent of ingested silicon is excreted in urine, the National Academies note.
Terms and sources
- the National Academies note
- National Academies, 2001: "Silicon in blood exists almost entirely as silicic acid and is not bound to proteins".
- excreted in urine
- National Academies, 2001: "as much as 50 percent of ingested silicon is excreted in the urine". They write that this suggests "some dietary forms of silicon are well absorbed".
A biological role of silicon in humans has not yet been identified, the National Academies write.
Terms and sources
- has not yet been identified
- National Academies, 2001: "A biological role of silicon in humans has not yet been identified". EFSA, 2004: "Silicon has not been shown to be essential for humans". The National Academies add that various connective tissues "contain most of the silicon present in the body".
Where silicon comes from
EFSA puts typical intake from food at 20–50 mg of silicon a day. Supplements use silicon compounds or plant extracts; the forms section below compares them.
Terms and sources
- 20–50 mg
- EFSA, 2004: "The estimated typical dietary intake (20-50 mg silicon/day) corresponds to 0.3-0.8 mg/kg body weight/day in a 60 kg person."
Daily amounts and the upper limit
The National Academies note: "There is no evidence that silicon that occurs naturally in food and water produces adverse health effects." It adds: "Limited reports indicate that magnesium trisilicate (6.5 mg of elemental silicon per tablet) when used as an antacid in large amounts for long periods (i.e., several years) may be associated with the development of urolithiasis due to the formation, in vivo, of silicon-containing stones (Haddad and Kouyoumdjian, 1986)." It continues: "Less than 30 cases of urolithiasis reported to be associated with intake of silicates (in the form of antacids) could be found even though antacids containing silicon have been sold since the 1930s."
Terms and sources
- The National Academies
- National Academies, 2001, page 531, section on hazard identification.
The same section adds: "Takizawa and coworkers (1988) examined the carcinogenicity of amorphous silica (SiO2) given by the oral route to rats and mice for approximately 2 years." It continues: "There was no evidence that orally administered silica induced tumors."
On the dose-response assessment, the National Academies write: "There are no adequate data demonstrating a no-observed-adverse-effect level (NOAEL) for silicon." They add: "Apart from scattered reports of silicate-induced urolithiasis, said to be associated with antacids, the limited toxicity data on silicon suggest that typical levels of intake have no risk of inducing adverse effects for the general population." They conclude: "Due to lack of data indicating adverse effects of silicon, it is not possible to establish a UL."
Terms and sources
- dose-response assessment
- National Academies, 2001, pages 531–532, section on dose-response assessment.
- UL
- Tolerable Upper Intake Level, a US upper limit. National Academies.
United States
No US recommended amount and no US upper limit are set for silicon. In US foods, silicon dioxide is also allowed as an anticaking agent, up to 2% of the food's weight.
Terms and sources
- No US recommended amount
- National Academies, 2001: "neither an Estimated Average Requirement, Recommended Dietary Allowance, nor Adequate Intake was established for silicon". On an upper limit: "it is not possible to establish a UL".
- anticaking agent
- 21 CFR 172.480(b)(4): not to exceed 2 percent by weight of the food. Paragraph (d) also allows it as an adsorbent in tableted foods for special dietary use. This covers only two vitamin ingredients, dl-alpha-tocopheryl acetate and pantothenyl alcohol. It is an additive, not an active ingredient. Our label count below covers only Supplement Facts rows.
The National Academies add, in a note to their table of upper limits: "Although silicon has not been shown to cause adverse effects in humans, there is no justification for adding silicon to supplements."
Terms and sources
- in a note to their table of upper limits
- National Academies, Dietary Reference Intakes, 2001, Summary, note to the table of upper limits. The same summary says: "Where data about adverse effects are extremely limited, extra caution may be warranted."
Average silicon intake in the Total Diet Study was 40 mg a day for men and 19 mg for women. The National Academies put the median intake of supplemental silicon by adults at about 2 mg a day.
Terms and sources
- Total Diet Study
- National Academies, 2001: "the mean intakes of silicon in adult men and women were 40 and 19 mg/day, respectively". The study is cited as Pennington, 1991.
- median intake
- National Academies, 2001: "The median intake of supplemental silicon by adults was approximately 2 mg/day". It cites the Third National Health and Nutrition Examination Survey.
In the NIH label database, 2,405 US labels list silicon in the Supplement Facts. Of them, 99% list more than one row in the Supplement Facts panel. 40% name silica or silicon dioxide as the form, 11% horsetail and 8% bamboo. 4.2% name choline-stabilized orthosilicic acid, 1.0% other orthosilicic acid and 0.3% MMST. 28% name no form.
Terms and sources
- 2,405 US labels
- NIH Dietary Supplement Label Database, including products no longer sold. Rows named silicon, silica, silicic or orthosilicic acid, ch-OSA, MMST or Silicea. Arginine, calcium, sodium and magnesium silicates were not counted. We counted these.
- 99%
- 2,378 of the 2,405 labels. We counted these.
- 40% name silica
- A label can name more than one form. Counted from the form field only; the row name is not counted as a form. Amino acid chelate 2.1%, magnesium trisilicate 1.9%, sodium metasilicate 1.9%, calcium silicate 1.7%. We matched whole words within one form name.
- 28% name no form
- 669 of the 2,405 labels. A label counts here when none of its silicon rows lists a form. We counted these.
Rows named Silicon list a median of 2 mg per serving. The middle half list 2–6 mg. 170 of the 1,263 rows list more than 10 mg; the highest list 140 mg.
Terms and sources
- Rows named Silicon
- 1,263 rows on these labels, with an amount. A serving is not always a full day's amount. We counted these.
- more than 10 mg
- The most common amount is 2 mg, on 473 rows. 7 labels list only a row named Silicon: 5 to 22.5 mg per serving.
Upper limits in other countries: see the table of limits by country on this page.
Published studies of silicon
4 studies: all 4 found a difference; in 3 of them the main measure was silicon in blood or urine.
Terms and sources
- 4 studies
- Healthy adults; from single doses to 20 weeks. From 10 to 750 mg of silicon a day; one study does not state its dose in the abstract. One study measured skin.
Studies in adults without a diagnosis, each with our one-line summary.
Not shown here: studies in people with a diagnosis or an illness, and studies in children. Studies that may include teenagers are not shown. Also not shown: studies that picked people by bone density. Silicon mixed with other ingredients, or given in beer, mineral water, spirulina or food, is left out too. So are products applied to the skin or mouth, and studies without a stated random order. Studies with a medicine arm are not shown. Arginine silicate products are left out: arginine is their main ingredient. Other reports and letters on listed studies are left out too.
Our one-line summary comes first. The original is on PubMed.
Show the studies
In 50 women, skin roughness fell with 10 mg silicon a day as ch-OSA and rose with a dummy.
Trial · 50 people · 20 weeks · industry-funded
Barel et al., Archives of Dermatological Research, 2005
Terms and sources
- 50 women
- Women with sun-damaged (photodamaged) facial skin. For 20 weeks, 25 took ch-OSA pellets with 10 mg of silicon a day and 25 took a dummy. Assignment was random and double-blind. Two corrections were published. One fixes the unit for silicon in Table 1: micrograms per liter, not per deciliter. The other fixes an author's name.
- skin roughness fell
- Forearm skin roughness, three measures: −16%, −19% and −8% with ch-OSA; +8%, +11% and +6% with the dummy. The difference between groups was statistically significant for two of the three (Rt and Rm). Blood silicon was higher with ch-OSA than with the dummy after 20 weeks. Hair and nail brittleness scores fell from the start with ch-OSA. The abstract compares them only with the start. The abstract does not report side effects. We read only the abstract.
- industry-funded
- A grant from Bio Minerals n.v. funded the study. The article says ch-OSA was developed for Bio Minerals n.v. by Dirk Vanden Berghe, a co-author. The article lists Vrije Universiteit Brussel and the University of Antwerp.
Study · Abstract only · about 2 min to read
Abstract only. Our estimate at about 200 words a minute, from the PubMed abstract.
Study summary
In 22 women, blood silicon was 272 mcg/L after 4 weeks of MMST and 191 with a dummy drink.
Trial · 14 women had complete blood data and 18 complete urine data. · 4 weeks on each, in turn · maker-funded
Jugdaohsingh et al., Nutrition & Metabolism, 2013
Terms and sources
- 22 women
- Healthy women aged 22–38 in London, UK, taking no medicines. Each took MMST with 10.5 mg of silicon a day for 4 weeks, and a dummy drink for 4 weeks. The order was random and double-blind. There was no wash-out between the two periods.
- 272 mcg/L
- Fasting blood silicon, means of the 14 women with complete blood data; 173 mcg/L at the start. Fasting urine silicon: 17.0 mg/L with MMST and 7.8 with the dummy, in 18 women. Both differences were statistically significant. Blood tests of liver and kidney function, blood fats and inflammation did not change with MMST. Nor did self-rated well-being.
- MMST
- Monomethylsilanetriol, taken as 3.5 mg of silicon three times a day. A letter by D. Vanden Berghe argued that 4 weeks in 22 women are too few to judge long-term use. The authors replied; they agreed that a formal assessment also needs animal toxicity data. Their reply states that he developed ch-OSA, another form of silicon, for Bio Minerals n.v.
- maker-funded
- The manufacturer, LLR-G5 Ltd (Ireland), supplied the MMST and the dummy drink and paid for one laboratory analysis. The authors’ laboratories have received research funding from the silicon supplement and food industry. Two authors have consulted for LLR-G5 and others in the silicon supplement industry.
Study · Free full text · about 14 min to read
Our estimate at about 200 words a minute, from the PMC full text.
Study summary
In 14 adults, blood silicon over 6 hours was higher after one dose of OSA-vanillin than after a dummy.
Trial · 14 people · Single doses, on two days · industry author
Marcowycz et al., Molecular Nutrition & Food Research, 2015
Terms and sources
- 14 adults
- Healthy men and women. On separate days, each took one dose of orthosilicic acid stabilized with vanillin (OSA-vanillin) and one dummy dose. The order was random and double-blind. The silicon dose is not stated in the abstract.
- was higher
- Blood silicon over 6 hours (area under the curve) was higher after OSA-vanillin. So was urine silicon over 6 hours. Both differences were statistically significant. The abstract gives no group values. About 21% of the silicon taken appeared in urine within 6 hours. The abstract does not report side effects. We read only the abstract.
- industry author
- Three of the six authors are from DexSil Labs (Belgium) and three from Biofortis (France). We could not check the funding section.
Study · Abstract only · about 1 min to read
Abstract only. Our estimate at about 200 words a minute, from the PubMed abstract.
Study summary
In 12 men, peak blood silicon was 2,855 and 2,498 mcg/L with two horsetail teas, 294 with low-silicon water.
Trial · 12 people · 3 days · funding not stated
Waterstradt et al., Planta Medica, 2021
Terms and sources
- 12 men
- Healthy male volunteers, after 36 hours of a low-silicon diet. Each arm drank 1 liter a day for 3 days. One tea held about 200 mg of silicon per liter, the other about 750 mg. The third arm drank low-silicon water. Arms were assigned at random; the abstract does not say how many men were in each.
- peak blood silicon
- Highest blood silicon after the first liter, at 60 minutes with the weaker tea and 120 minutes with the stronger. At 120 minutes, blood silicon differed significantly between the stronger tea and water. Silicon did not build up over 3 days. Urine silicon over 24 hours: 109,000 mcg (weaker tea), 142,000 (stronger) and 19,000 (water). Immune measures did not change significantly.
- two horsetail teas
- Tea of Equisetum arvense (field horsetail). The authors call the weaker tea TEA1 and the stronger one TEA2. The authors report: "TEA1 was well tolerated, while TEA2 caused diarrhoea in 4 subjects." The abstract reports no side effects for water.
- funding not stated
- We could not check the funding section. PubMed lists the University of Freiburg (Germany) and the University of Basel (Switzerland). The trial is registered as DRKS00016628.
Study · Abstract only · about 2 min to read
Abstract only. Our estimate at about 200 words a minute, from the PubMed abstract.
Study summary
Information, not a prescription. Do not start, stop or combine supplements or medicines on your own.