Colloidal gold: the physics that fails, and the chemistry that might
Gold sounds like a scam the moment someone says it sharpens the mind, calms the joints, and makes life move faster. Most of the reasons people give are real properties of the metal used at the wrong scale. Gold is diamagnetic. It is the best electrical conductor we have. A red colloid is red because the particles swallow green light. None of that reaches your life from a spoonful. What can reach you is slower and stranger: a particle your immune cells eat, dissolve, and turn into a trickle of gold ions inside the exact cells that run inflammation.
- 2 mlOf a 180 ppm bottle, matching the only human joint trial
- ~2%Of swallowed gold particles that show up in blood, in rats
- WeeksBefore the stored-gold effect can be read
- HoursIf a same-day effect exists, it is happening in the gut
The properties that do not scale
Diamagnetic means a substance is weakly pushed out of a magnetic field. Water is diamagnetic too, and you are mostly water. The difference between gold and water in Earth’s field is so small that the heat already jiggling a 20 nanometre particle is tens of billions of times stronger than the magnetic push. The heart’s magnetic field is about half a million times weaker than Earth’s. The brain’s is weaker again. Even a hospital MRI magnet does not put a free gold particle under magnetic control. Nothing in you is steering these particles with magnetism.
Conductivity fails for a different reason. Each particle is a superb conductor sitting in salt water that already conducts, then immediately coated in your own proteins. That coat insulates it. Nerves do not speed up because the fluid around them got a better metal. They run on ion channels, and the fast ones are wrapped in myelin, which is an insulator. Gold conducts in the body when it is a wire with a battery on the other end. A loose particle has no wire.
The red colour is a real optical trick. Particles around 10–30 nanometres across absorb green light near 520 nanometres and turn it into heat. That only matters if a strong light hits them. Green light dies within about a millimetre of skin. A desk lamp is thousands of times too dim, and the wrong distance, compared with the lasers used when researchers actually heat gold particles on purpose. Gold is a converter. It needs a hard input. The body’s own electrical buzz is the wrong frequency by an absurd margin, and far too weak.
Catalysis is the one physical property that almost survives. Clean gold surfaces can break down hydrogen peroxide the way an enzyme does. They can also oxidise glucose and make peroxide, so the same surface cuts both ways. In blood, the protein coat covers those surfaces and most of the activity is gone. The marketing version of “gold is a catalyst in your blood” skips the coat.
The mechanism that fits a small dose and a large one
Particles that cross the gut are eaten by macrophages, the immune cells parked in the gut wall, the liver, the spleen, and any tissue that is already inflamed. Inside the cell’s acid compartment, gold slowly dissolves. Cells have been watched breaking gold nanoparticles down and rebuilding the metal as tiny crystals. The gold that comes off is an ion. That ion binds selenium in thioredoxin reductase, the same target as the old arthritis drug auranofin. With that enzyme damped, the inflammatory switch NF-κB drives less interleukin-1, interleukin-6, and TNF.
One switch, many complaints. Those cytokines suppress the brain signal that tells the testes to make testosterone. They push hair follicles toward shedding. They keep gums puffy and thinking foggy. Lower the signal and all of those can ease together, which is why one substance picks up a reputation for everything. It does not block DHT, so a male-pattern hairline does not redraw itself. It does not become bone. Gold is not a mineral the skeleton lays down.
There is a faster door, and it does not need the particle to reach the blood. The entire swallowed dose washes past the immune cells in the gut at a much higher concentration than anything in the bloodstream. If those cells turn down their response to bacterial residue, the “sick, foggy, heavy” signal up the vagus nerve drops. That can be felt the same day. The slow store in macrophages is felt over weeks and can linger after you stop, because the metal is still sitting there.
Both can be true in the same person. A same-day sharpness and a change that outlasts the bottle are two clocks, not a contradiction. The first is the gut. The second is the store.
The overlooked interaction: thyroid
The enzymes that activate and destroy thyroid hormone are selenium enzymes, the deiodinases. Gold ions bind that selenium. In liver microsomes, gold salts and auranofin inhibit the main one, deiodinase 1, at well under a micromolar. Deiodinase 2 and 3 are less sensitive. Deiodinase 1 clears used-up thyroid hormone in the liver. Deiodinase 3 inactivates T3 in the brain.
Someone taking T3 directly, with their own thyroid output already suppressed, does not need these enzymes to make hormone. What a gold ion would change is how fast that T3 is broken down. Slower breakdown is a quiet dose increase. It would feel like a bit more drive, warmer hands, a slightly faster pulse, more appetite. A lab test of free T3, on a fixed number of drops, is the clean way to see it. If free T3 rises on gold and falls off it, the thyroid route is real, and the drops need to come down to match.
Particle size matters here. In that same enzyme assay, 5 nanometre particles inhibited deiodinase 1, and 30 nanometre particles did not. A red colloid is mostly the larger size. The thyroid route probably needs dissolved ions, meaning weeks of a real dose, not a sip.
Why life can feel like it speeds up
A little more effective thyroid hormone, before it turns unpleasant, feels like sharper thinking, more nerve, less patience for waiting, and a sense that things are lining up. That feeling has a boring route into the real world. You act sooner, you ask, you follow up, and people answer someone who looks like they mean it. Hits get remembered. Misses fade. The metal’s interaction with green light does not reach events outside your skin. The path from the molecule to a faster week runs through behaviour, and it is still a real path.
The failure mode is the same pathway with the gain too high. Short sleep that does not feel like short sleep, thoughts you cannot steer, spending or risk that shows up overnight, irritation when anyone slows you down. Thyroid hormone above replacement, sleep debt, and heavy cannabis all push that direction. Gold only joins them if it is truly slowing T3 breakdown. If those signs show up, the gold comes out first and the thyroid dose comes back to replacement.
Dose, in swallowed gold and in gold that arrives
Bar length: estimated gold reaching the blood, micrograms per day, assuming about 2% of swallowed nanoparticles are absorbed, from rat data. Auranofin is plotted as absorbed drug, not as nanoparticles. This is a scale, not a blood test. The amber mark is auranofin absorbed, 400 μg a day.
- Knee trial, 0.34 mg 6 μg
- 2 ml of 180 ppm 6 μg
- 15 ml of 180 ppm 50 μg
- 5 mg a day 90 μg
- 20 mg a day 370 μg
- Auranofin, 6 mg 400 μg
180 ppm is 2.7 mg of elemental gold per 15 ml. The knee trial used 0.34 mg a day of 8–28 nm particles for 20 weeks. Rat oral absorption was about 1.85% for particles and about 8.5% for gold ions. Auranofin is 6 mg of compound, 29% gold, absorbed far better than a particle because it was built that way.
How to read that chart
The only human trial of oral gold nanoparticles used a small dose and got a soft result: people with bad knees said the knee was better after 20 weeks, and most hard measures only drifted. Matching that swallowed dose is about 2 ml of a 180 ppm bottle, not the 15 ml listed as a serving. Five milligrams a day of particles is still well under the absorbed dose of the old gold pill. Twenty milligrams a day is in the same absorbed neighbourhood as auranofin, which means it is no longer a trace experiment. At that range the old drug caused rash, mouth ulcers, and, less often, protein in the urine or a drop in blood counts.
A red bottle is particles. If a product is yellow and clear, more of it may be ionic gold, which is absorbed several times better. The label’s milligrams then understate what reaches the blood. Colour is a clue, not an assay.
What the famous claims actually were
| Claim | What was actually done | What it can support |
|---|---|---|
| Joints | 0.34 mg elemental gold a day, 20 weeks, 51 adults | A possible slow effect on an inflamed joint. Subjective scores moved. Most objective tests did not clearly. |
| Hair | A plant extract stuck to gold particles, rubbed on mice | Not oral gold. The plant was doing the work. Gold does not block DHT. |
| Bone | No oral gold trial of bone density worth using | Nothing. Gold is not bone mineral. Calmer inflammation is not a denser skeleton. |
| IQ | Five people, 30 mg a day, four weeks, no placebo, testers not blinded | Not a dose to copy. Scores rose and later fell. The gold pill at a similar absorbed dose never produced this in decades of use. |
| Testosterone, mood, gums | No direct trial at colloidal doses | One mechanism: less inflammatory brake. Weeks, not an afternoon, and only if that brake was on. |
A way to test it that isn’t a testimonial
Have someone else dose a dark drink, gold on some days and none on others, in random order, ten of each. You don’t get told which. Two or three hours later, write a sharpness score, run a two-minute reaction test, take a resting pulse, and note whether the hands are warm. Gold days that feel sharper with no change in pulse point at the gut. Pulse and warmth rising too point at thyroid. No difference on any single day, but a change after three weeks, points at the store, and that needs on-and-off blocks of weeks, not a morning guess.
If you already take T3, fix the drops first and don’t move them during the test. A morning free-T3 blood test at the end of three weeks off and three weeks on is the result that would actually be new. Free T3 up, drops unchanged, means gold is slowing the breakdown and the drops should be lowered to compensate.
Sources worth reading beside this: the 20-week oral gold nanoparticle knee study at 0.34 mg a day; rat oral biokinetics showing about 1.85% absorption for gold nanoparticles and about 8.5% for gold ions; the 2021 finding that gold ions and 5 nm particles inhibit human deiodinase 1; auranofin’s label, 6 mg a day, 29% gold. The 1998 five-person IQ pilot is included as a warning about dose, not as evidence. This piece is a mechanism argument. It is not a recommendation to drink a gold colloid, and it is not a substitute for a blood test if you already take thyroid hormone.