Home › Supplement Vault › Silica

Silica

Best-in-class: Silica

Skin & Structural✅ Clinically validated📊 Correlative data🧪 Theoretical

A mineral essential for collagen cross-linking and connective tissue — supports hair strength, nail growth, skin and bone.

Educational use only — not medical advice. These statements have not been evaluated by the FDA. This product is not intended to diagnose, treat, cure or prevent any disease.

Silica quick facts

Suggested doseAs directed (choline-stabilized orthosilicic acid or bamboo extract).
How oftenDaily
Who it's forHair, skin, nail and bone/connective-tissue support.
Coach Cam’s take

A genuine structural cofactor with a real mechanism, and thin human repletion data. Small trials show improved skin, hair and nail quality. The form matters enormously for absorption — most cheap silica supplements deliver very little bioavailable silicon. Horsetail contains thiaminase, which degrades vitamin B1, so long-term use of that source is not ideal.

How Silica actually works

Orthosilicic acid is the bioavailable form, and silicon is required for the cross-linking of collagen type I and for early bone matrix mineralization — deficiency produces defective connective tissue in animal models. Bamboo and horsetail supply it as silica that must be converted, which is inefficient; choline-stabilized orthosilicic acid is the pre-formed version.

⚠️ Good to know: Use a bioavailable form (orthosilicic acid) — plain silica is poorly absorbed.

Where to get Silica

Find Silica on iHerb →
Top-rated brands on iHerb · Coach Cam partner link

The evidence for Silica

Graded by what exists behind each claim.

✅ Clinically validated

📊 Correlative data

🧪 Theoretical / extrapolated benefits

How to read these tiers: they say how much human evidence exists, not how well something works — and ✗ flags harm, never a disappointing trial. How the evidence tiers work →

What Silica actually does

There is exactly one chemical species of silicon that a human body can use, and almost nothing sold as silica is it. The absorbed and circulating form is orthosilicic acid, Si(OH)4 — a small, uncharged, water-soluble molecule. Above low concentrations orthosilicic acid polymerizes: molecules condense into oligomers and then into colloidal and finally solid amorphous silica, and polymerized silica is not absorbed. Every formulation question on this page is a question about keeping silicon monomeric.

The cell experiment is specific and it has concentrations. Orthosilicic acid applied to human osteoblast-like cells at 10, 20 and 50 micromolar stimulated type 1 collagen synthesis and osteoblastic differentiation Reffitt 2003. Note which way round that is: the primary effect is on the collagen — the protein scaffold — and on the differentiation state of the cell making it, not on mineral deposition. Bone is roughly a third collagen by weight, and a bone with defective collagen is brittle regardless of how much calcium is in it.

That is also the correct mechanistic account of the hair, skin and nail claims, and it is more modest than the marketing. Hair and nail are keratin, not collagen; the dermis underneath skin is collagen. So a silicon effect on collagen synthesis Reffitt 2003 predicts changes in the dermal matrix and in the tissue anchoring the hair follicle and nail bed — mechanical properties like elasticity and break strength — rather than a change in how fast hair or nail grows. The card's own theoretical note says exactly this, and it is right.

The population arithmetic, so the doses have context. Habitual dietary silicon intake in a Western population is 30 to 33 mg a day in men and 24 to 25 mg a day in women (P = 0.0001), it decreases with age (P < 0.001, adjusted for sex), and a mean of 41% of ingested silicon is excreted in the urine Jugdaohsingh 2002. Urinary excretion is how absorption is measured for this element, because silicon is not metabolized — what is absorbed is cleared by the kidney. A supplement delivering 10 mg of silicon is therefore adding roughly a third of the dietary intake, not multiplying it.

And the epidemiology points at the same tissue, with one restriction that matters enormously. In the Framingham Offspring cohort, dietary silicon intake was positively associated with bone mineral density, with up to a 10% difference in hip BMD between the highest quintile (over 40 mg Si/day) and the lowest (under 14 mg Si/day) — in men and premenopausal women Jugdaohsingh 2004. The association is stated for those two groups. Postmenopausal women, the group most likely to buy a silicon supplement for bone, are the group the finding does not describe.

Cell, rodent, human — and where it stops

Cell to human is short here, and the human absorption study is the most useful single table in this cohort. Silicon absorption was measured across sources, as a percentage of the ingested dose recovered in urine: monomethyl silanetriol and alcohol-free beer 64%; green beans 44%; orthosilicic acid 43%; choline-stabilized orthosilicic acid 17%; bananas and magnesium trisilicate antacid 4%; colloidal silica 1% Sripanyakorn 2009.

Read that list twice, because it inverts the shopping advice. Colloidal silica — which is what bulk ‘silica’ and most food-grade silicon dioxide is — absorbs at one percent, and the European re-evaluation of silicon dioxide as a food additive independently concluded that it is poorly absorbed Younes 2018. That part everyone gets right. But choline-stabilized orthosilicic acid, the branded ‘bioavailable’ form, absorbed at 17% — four times worse than plain orthosilicic acid at 43% and a quarter of monomethyl silanetriol at 64% Sripanyakorn 2009. The form with all the clinical trials is not the form with the best absorption.

The trials belong to the 17% form, and they are real. Skin: 50 women took 10 mg Si/day as ch-OSA for 20 weeks; skin roughness rose in the placebo group (Rt +8%, Rm +11%, Rz +6%) and fell in the ch-OSA group (Rt −16%, Rm −19%, Rz −8%), with nail and hair brittleness scores significantly lower Barel 2005. Hair: 48 women with fine hair took the same 10 mg Si/day for nine months; the elastic gradient fell only 4.52% against 11.9% on placebo, and break load fell 2.20% against 10.8% Wickett 2007. Bone: of 184 women, 136 completed 12 months on 1,000 mg calcium and 20 mcg cholecalciferol plus ch-OSA at 3, 6 or 12 mg Si or placebo; the type 1 collagen formation marker PINP was significantly higher at 12 months for the 6 and 12 mg arms Spector 2008.

Notice what those three trials have in common, because it is the argument. Every endpoint that moved is a mechanical or collagen-formation endpoint — roughness, elasticity, break load, PINP. That is exactly what a compound stimulating type 1 collagen synthesis in osteoblasts at 10–50 microM should do Reffitt 2003. The mechanism and the outcomes agree, which is unusual and worth crediting.

So the specific obstacle is that nobody has measured the thing in the middle. There is a cell effect at 10–50 microM Reffitt 2003 and there are clinical endpoints at 10 mg Si/day Barel 2005 Wickett 2007, and no trial reports the serum silicon concentration those doses produce. Absorption is measured by urinary recovery Sripanyakorn 2009 Jugdaohsingh 2002, not by a steady-state plasma level held against the concentration the osteoblasts needed. Until somebody publishes that number, the chain from the dish to the fingernail has a missing link in the middle.

The second obstacle is who the trials were in. All three ch-OSA trials were in women, selected for photodamaged skin, fine hair, or osteopenia Barel 2005 Wickett 2007 Spector 2008, and the bone trial gave calcium and vitamin D to everybody. These are deficiency-shaped or deficit-shaped populations, and the epidemiological association is reported for men and premenopausal women Jugdaohsingh 2004. Nobody has shown a silicon effect in somebody with normal hair, normal skin and a normal DXA.

Silica — which form, and does it matter

The card's own form field is wrong, and correcting it is the most useful thing on this page. It reads ‘Bamboo/horsetail (orthosilicic acid)’. Bamboo and horsetail do not contain orthosilicic acid. Plants take silicon up as orthosilicic acid and then deposit it as solid amorphous silica — phytoliths — which is why horsetail is abrasive enough to have been used as a scouring rush. What a bamboo or horsetail capsule delivers is polymerized silica, and polymerized silica is the 1% and 4% end of the absorption table, not the 43% end Sripanyakorn 2009 Younes 2018.

Which is also why choline-stabilized orthosilicic acid exists at all. The filed label describes it plainly: silicon 5 mg as choline-stabilized orthosilicic acid plus choline 100 mg per capsule, one capsule twice daily Office of Dietary Supplements — 10 mg Si/day, exactly the trial dose Barel 2005 Wickett 2007. The choline is there to hold the silicic acid monomeric in a concentrated liquid, which is a real engineering solution to the polymerization problem, and it is why this is the form with the trials.

And yet it is not the best-absorbed form, which is the genuinely surprising finding. ch-OSA absorbed at 17% against 43% for plain orthosilicic acid and 64% for monomethyl silanetriol Sripanyakorn 2009. So the market has settled on the form with the clinical data rather than the form with the best absorption, and no head-to-head trial has tested whether the better-absorbed forms produce the same skin and hair endpoints. That is an evidence artifact, not a pharmacological conclusion, and buyers should know which one they are paying for.

Two entries in that table are worth pulling out for anyone who does not want to buy a supplement. Alcohol-free beer absorbed at 64% and green beans at 44% Sripanyakorn 2009. Beer is a well-known dietary silicon source because barley is a silicon accumulator and the brewing process leaves the silicon in solution as orthosilicic acid rather than as phytolith. A person whose habitual intake is at the top Framingham quintile of over 40 mg Si/day Jugdaohsingh 2004 got there from food, not from a capsule.

What a label should carry. Milligrams of elemental silicon, not milligrams of the source material — a 500 mg horsetail capsule tells you the weight of the plant. And the chemical form, because the difference between the top and bottom of the absorption table is sixty-four-fold Sripanyakorn 2009. Almost no product on the shelf states either.

What would have to be true, and how you would know it was not

1. Nail break strength and hair elasticity, with the trial's own effect sizes and the honest direction. On 10 mg Si/day as ch-OSA, predict that over nine months hair elastic gradient declines by about 4.5% rather than the 11.9% it would otherwise, and break load by about 2.2% rather than 10.8% Wickett 2007. Read that carefully: in the trial the treated group still got worse, just less. The honest prediction is slowed deterioration, not improvement, and a product sold on regrowth is selling something its own evidence did not show.

2. Timing, because it is set by biology and not by marketing. The skin trial ran 20 weeks Barel 2005, the hair trial nine months Wickett 2007, the bone trial twelve months Spector 2008. A fingernail takes months to grow out from the fold. Predict nothing detectable before four months, and mark the nail at the base with a fine file at the start so the before-and-after boundary is physically visible.

3. p1np, the marker that turns this into a measurement. Procollagen type 1 N-terminal propeptide is a routine bone-formation assay and it rose significantly at 12 months on 6 and 12 mg Si/day — but not on 3 mg Spector 2008. So predict P1NP up at 6 mg or more and flat at 3 mg. That is a dose threshold from a real trial, it makes the ‘which dose’ question answerable, and P1NP is the same collagen-formation readout the cell experiment predicts Reffitt 2003.

4. Bone mineral density — the prediction that cuts against the product. Predict no detectable DXA change at 12 months. The bone trial's positive result was a formation marker, on top of calcium and vitamin D given to everyone Spector 2008, and the epidemiological association is reported in men and premenopausal women Jugdaohsingh 2004 — not in the group most likely to be taking it for osteoporosis. A rising P1NP with an unchanged DXA is the expected result and it is a real finding, not a failure; it just is not the one on the box.

5. The experiment that would make this page obsolete, and it is one blood draw. Take your product, and have serum silicon measured before and two hours after a dose. Predict that a colloidal or horsetail product moves it barely at all — 1% and 4% absorption Sripanyakorn 2009 — and that an orthosilicic acid product moves it measurably. No published trial reports serum silicon alongside a hair, skin or bone endpoint, so this is the cheapest unclaimed measurement in the category, and it would tell you in an afternoon whether the capsule you bought delivers anything at all.

What nobody has tested yet

Nobody has published the serum silicon concentration that a clinical dose produces. The cell effect needs 10 to 50 micromolar Reffitt 2003; absorption is quantified by urinary recovery Sripanyakorn 2009 Jugdaohsingh 2002; and the trials report skin, hair and bone markers without a plasma concentration Barel 2005 Wickett 2007 Spector 2008. The single measurement that would connect the mechanism to the outcome has never been reported.

Nobody has run the better-absorbed forms to a clinical endpoint. Monomethyl silanetriol absorbs at 64% and plain orthosilicic acid at 43%, against 17% for the form with all the trials Sripanyakorn 2009. A three-arm study at matched elemental silicon doses, with P1NP and hair break load as endpoints, would tell the field whether the market has been buying the right molecule for twenty years.

Nobody has tested silicon in men or in postmenopausal women for hair, skin or bone. All three ch-OSA trials were in women selected for a deficit Barel 2005 Wickett 2007 Spector 2008, and the epidemiological association is stated for men and premenopausal women Jugdaohsingh 2004. The two populations that buy this most — men with thinning hair and postmenopausal women with low bone density — are the two with no interventional data.

And nobody has measured what horsetail actually delivers. The absorption table has an entry for colloidal silica at 1% Sripanyakorn 2009 and none for a horsetail or bamboo extract. Given that plant silicon is deposited as phytolith, the expected answer is ‘close to nothing’, but nobody has fed a standard horsetail capsule to volunteers and measured urinary silicon. It is a one-day study and it would settle the largest segment of this market.

Silica — its own safety story, not its category's

The first thing to separate is the thing people confuse this with. Occupational exposure to respirable crystalline silica dust causes silicosis and is associated with lung cancer mortality in a dose-dependent way Checkoway 1997. That is an inhaled, crystalline, industrial exposure. Dietary silicon is ingested, amorphous or dissolved, and the European additive re-evaluation found silicon dioxide poorly absorbed, with no genotoxicity concern and dietary exposure estimates at least an order of magnitude below the no-observed-adverse-effect levels identified Younes 2018. Nothing on this page is about lungs.

Kidney disease is the one real contraindication and the mechanism is straightforward. Silicon is not metabolized; what is absorbed is excreted renally, and 41% of ingested silicon appears in urine Jugdaohsingh 2002. Remove the clearance route and the absorbed fraction accumulates. Anyone with impaired kidney function is taking a mineral whose only exit is the organ that is not working, and that argues against supplementation rather than for a dose adjustment.

Horsetail carries a thiaminase caution, and the toxicology is more reassuring than the caution implies. Equisetum arvense fed to F344 rats at 0, 0.3, 1 and 3% of the diet for 13 weeks produced no toxicity in clinical signs, body weight, urinalysis, hematology, serum biochemistry or organ weights, and the no-observed-adverse-effect level was set at greater than 3% of diet — more than 1.79 to 1.85 g/kg body weight per day Tago 2010. That is a large exposure with a clean result. The thiaminase concern is real in principle, comes from veterinary poisoning in grazing animals eating horsetail as forage, and is not supported at supplement doses by this study. Taking a B-complex alongside a horsetail product costs nothing and closes the question.

The choline in a choline-stabilized product is worth counting. The filed label carries 100 mg of choline per capsule alongside 5 mg of silicon, at two capsules a day Office of Dietary Supplements — 200 mg of choline nobody mentions. That is a modest amount and it is not a hazard; it is simply an ingredient the buyer did not know they were taking, and it matters to anyone counting choline for another reason.

And the honest closing point is about expectation rather than harm. This is among the safest things in the catalog: the controlled trials at 10 mg Si/day ran 20 weeks, nine months and twelve months without a safety signal Barel 2005 Wickett 2007 Spector 2008. The cost is not toxicity. It is that the trials showed slowed deterioration in mechanical properties of hair and skin and a bone formation marker, not regrowth and not a density change — and that a horsetail or bamboo capsule, which is what most people actually buy, delivers a form that absorbs at a few percent Sripanyakorn 2009 Younes 2018.

Sources read for this page

How you would know if it worked

There is no blood test for this one. That is not a criticism — it is a fact about the effect, and it changes how you should judge it.

Run it one variable at a time. Starting three things in one week means a result you cannot attribute, which is the same as no result.

Silica — safety & side effects

Not medical advice. If you take prescription medication or have a diagnosed condition, check this against it with a pharmacist or doctor — pharmacists are underused and free.

🔒
The dose is the easy part. Making Silica actually work is what's behind Skool:
Running it
  • When to take it, and what to take it with
  • Which form actually absorbs
  • Who it's worth it for
  • Best-in-class brand pick
  • Coach Cam's stacks and notes
When to take it
  • Fasted or with food, and when in the day
  • Morning or night, and why that window
  • Around training, or deliberately away from it
  • What it must not share a window with

Everything above is free and stays free. Skool is where it becomes a plan — Silica in an order, with the rest of what you're running.

Unlock in Skool — $10/mo →

Bloodwork to run alongside Silica

Baseline first, then again at 8–12 weeks.

MarkerWhat it’s watching for
FerritinBelow 50 and hair and skin repair suffer, whatever else you take
Vitamin D (25-Hydroxy)Skin, bone and connective tissue all depend on it
TSH (Thyroid-Stimulating Hormone)Thyroid disease shows in hair, skin and nails first
Zinc, PlasmaDeficiency causes poor wound healing and hair shedding

The Basics — Start Here panel covers these in one order — 4 markers, $32.40 with the discount applied.

Check results you already have → · All 103 markers A–Z

Silica — frequently asked questions

What is Silica?

A mineral essential for collagen cross-linking and connective tissue — supports hair strength, nail growth, skin and bone.

What is the suggested dose of Silica?

As directed (choline-stabilized orthosilicic acid or bamboo extract). This is a general reference for education only — statements have not been evaluated by the FDA and this is not medical advice.

Where can I find Silica dosing and the full breakdown?

The suggested dose and the full evidence — clinical, correlative and theoretical — are on this page. What's inside Skool is when to take it, which form actually absorbs, the brand worth buying and Coach Cam's stacks.

Where can I buy Silica?

Coach Cam sources Silica from vetted, top-rated brands on iHerb — use the buy link on this page.

What Silica is used for

Silica appears under 3 goals in the goal router.

🩹 Heal an injuryCollagen & matrix synthesis🩹 Heal an injuryBone & fracture healing✨ Skin, hair & aestheticsCollagen synthesis & dermal matrix🦴 Joints & boneCartilage matrix & joint substrate🦴 Joints & boneBone remodeling — building vs preserving
CollagenGlucosamine & ChondroitinSkin ComplexBone SupportHyaluronic AcidMSMPhytoceramidesSea Buckthorn

Where this goes next

In contextCollagen & matrix synthesisThe heal an injury pathway Silica belongs to
Go deeper$10/mo

The pages here are the frameworks. The protocols — the dosing, the order to correct things in, the week-by-week schedule and what to retest — are inside Skool.

Every protocol on the site

← Browse the full Supplement Vault

↑ Back to on this page