Place · Level 3
Milk Thistle · Silybum marianum
全球卖得最好的护肝草药 · 试管里的机制是真的, 但口服几乎到不了肝 · 静脉那一版是鹅膏菌中毒的真解毒剂, 干的却是另一件事 · Cochrane 与 SyNCH 大试验阴性 · 安全, 但它替掉的东西才是代价
Story path
- 1What is actually in the capsuleWhat is actually in the capsule
- 2In a dish, it really does thingsIn a dish, it really does things
- 3The wall it has to get pastThe wall it has to get past
- 4The IV version is a different drugThe IV version is a different drug
- 5It has been tried in peopleIt has been tried in people
- 6A lower ALT is not a better liverA lower ALT is not a better liver
- 7Safe — but not the thing that worksSafe — but not the thing that works
Chapter 1
What is actually in the capsule
What is actually in the capsule
That bottle of 'liver support' pills says silymarin on the label. Silymarin is not a single compound — it is a mixture of structurally related molecules. That sounds like a footnote, but it determines every argument about dose that follows.
The plant
Milk thistle is Silybum marianum, in the daisy family (Asteraceae). Native to the Mediterranean, now widely cultivated. The white veining on its leaves gave it the 'milk' name.
The part used is the seed — not the leaf, not the flower.
What comes out of the seed
Purified seed extract is called silymarin, a collective name for a group of flavonolignans, at least four or five distinct molecules:
Silybin A and B (also written silibinin) — the most abundant and by far the most studied, roughly half the mixtureIsosilybinSilychristinSilydianinPlus some taxifolin
So the number on the box has to be read twice. A capsule labelled '300 mg extract, standardized to 80% silymarin' contains 240 mg of the mixture; the fraction that is actually silybin is only a bit over a hundred milligrams. Some papers dose by silymarin, others by silybin, and the two differ roughly two-fold. When two brands' doses don't line up, check which one they are counting.
This island answers three questions: what it does inside a cell, whether a swallowed capsule ever gets there, and why an emergency room genuinely does use it.
The plant
Milk thistle is Silybum marianum, in the daisy family (Asteraceae). Native to the Mediterranean, now widely cultivated. The white veining on its leaves gave it the 'milk' name.
The part used is the seed — not the leaf, not the flower.
What comes out of the seed
Purified seed extract is called silymarin, a collective name for a group of flavonolignans, at least four or five distinct molecules:
Silybin A and B (also written silibinin) — the most abundant and by far the most studied, roughly half the mixtureIsosilybinSilychristinSilydianinPlus some taxifolin
So the number on the box has to be read twice. A capsule labelled '300 mg extract, standardized to 80% silymarin' contains 240 mg of the mixture; the fraction that is actually silybin is only a bit over a hundred milligrams. Some papers dose by silymarin, others by silybin, and the two differ roughly two-fold. When two brands' doses don't line up, check which one they are counting.
This island answers three questions: what it does inside a cell, whether a swallowed capsule ever gets there, and why an emergency room genuinely does use it.
Chapter 2
In a dish, it really does things
In a dish, it really does things
Let's be fair first: silybin is not placebo powder. Add it to hepatocytes in culture and three things genuinely happen. Hold on to these three, because the next scene asks whether the dose you swallow can make any of them happen.
One · it crowds the doorway
Silybin inserts itself into the hepatocyte's outer membrane. That membrane carries transport proteins that haul molecules inward; silybin occupies space there, and some toxins become harder to haul in. This one matters enormously in the next scene.
Two · it helps absorb oxidation
When the liver dismantles alcohol, drugs, and foreign molecules, it throws off reactive oxygen — sparks from a workshop in operation. Cells catch those sparks with glutathione. Silybin scavenges some radicals directly, and in animal work it partially restores depleted glutathione. For how that system runs, see the glutathione and nac islands.
Three · it quiets the scarring cells
The liver holds a cell type called stellate cells, which normally sit quietly storing vitamin A. Under repeated injury they wake up and turn into collagen factories, plastering the damage layer by layer — that is fibrosis, and enough of it is cirrhosis. In culture, silybin dampens that wake-up signal.
Three coherent mechanisms, three with experimental support. Only one problem remains, and it is a physical one: at the concentrations those experiments used, will your hepatocytes ever see it?
One · it crowds the doorway
Silybin inserts itself into the hepatocyte's outer membrane. That membrane carries transport proteins that haul molecules inward; silybin occupies space there, and some toxins become harder to haul in. This one matters enormously in the next scene.
Two · it helps absorb oxidation
When the liver dismantles alcohol, drugs, and foreign molecules, it throws off reactive oxygen — sparks from a workshop in operation. Cells catch those sparks with glutathione. Silybin scavenges some radicals directly, and in animal work it partially restores depleted glutathione. For how that system runs, see the glutathione and nac islands.
Three · it quiets the scarring cells
The liver holds a cell type called stellate cells, which normally sit quietly storing vitamin A. Under repeated injury they wake up and turn into collagen factories, plastering the damage layer by layer — that is fibrosis, and enough of it is cirrhosis. In culture, silybin dampens that wake-up signal.
Three coherent mechanisms, three with experimental support. Only one problem remains, and it is a physical one: at the concentrations those experiments used, will your hepatocytes ever see it?
Depth · the stellate-cell step
Zoom in on the fibrosis step, because it is the most valuable claim in the entire liver-supplement market.A narrow space sits between hepatocytes, and stellate cells live in it. In health their job is quiet: store vitamin A, maintain a little matrix.
Sustained injury — alcohol, fat, virus, drugs — makes neighbouring macrophages release a signal called TGF-beta. Stellate cells receiving it transform: they dump their vitamin A stores, grow contractile fibres, and start mass-producing collagen. That is scar.
The crucial question is whether scar is permanent. It used to be assumed so; we now know it is not entirely — remove the source of injury and early fibrosis can partly regress. That is exactly why quitting alcohol, losing weight, and antiviral treatment are worth far more than any supplement: they act on the source, while a supplement at best negotiates downstream.
In vitro, silybin weakens that TGF-beta step. Whether it does so in a person is the clinical-trial scene later on.
Chapter 3
The wall it has to get past
The wall it has to get past
'Works in a dish doesn't mean works in a body' is usually a throwaway line. For milk thistle it has a specific, measurable reason: most of the silybin you swallow never reaches the inside of a hepatocyte.
Barrier one · it barely dissolves
Silybin is nearly water-insoluble, and the gut absorbs a molecule only after it disperses into the aqueous phase. What doesn't disperse leaves in stool.
Barrier two · it gets tagged on arrival
Whatever is absorbed travels the portal vein straight to the liver. Hepatocytes run a workshop called phase II enzymes, whose job is to hang a large water-loving chemical group (glucuronide or sulfate) onto foreign molecules. Once tagged, a molecule can no longer cross membranes and is flushed into bile or urine.
That workshop exists to protect you — it is how most foreign molecules are cleared. But for a molecule that wants to work inside hepatocytes, it is a wall: the silybin measurable in your blood is overwhelmingly the tagged conjugate, not the free active form. And clearance is fast, with a half-life of only a few hours.
What the wall looks like when measured
Someone measured it properly. A phase I study dosed hepatitis C patients every 8 hours, escalating from 140 mg to 700 mg. A 5-fold increase in dose produced a 38-fold increase in silybin B exposure (Hawke 2010).
That is counterintuitive, but it means one thing: at customary doses the clearance pathway is nowhere near saturated, so almost everything is intercepted. Only by pushing far above customary doses does the pathway start to saturate and blood levels begin to hold.
So when you next see an in-vitro claim, ask one question first: is that concentration reachable by swallowing? The question applies to every polyphenol supplement — curcumin is the same story, see the curcumin island.
Barrier one · it barely dissolves
Silybin is nearly water-insoluble, and the gut absorbs a molecule only after it disperses into the aqueous phase. What doesn't disperse leaves in stool.
Barrier two · it gets tagged on arrival
Whatever is absorbed travels the portal vein straight to the liver. Hepatocytes run a workshop called phase II enzymes, whose job is to hang a large water-loving chemical group (glucuronide or sulfate) onto foreign molecules. Once tagged, a molecule can no longer cross membranes and is flushed into bile or urine.
That workshop exists to protect you — it is how most foreign molecules are cleared. But for a molecule that wants to work inside hepatocytes, it is a wall: the silybin measurable in your blood is overwhelmingly the tagged conjugate, not the free active form. And clearance is fast, with a half-life of only a few hours.
What the wall looks like when measured
Someone measured it properly. A phase I study dosed hepatitis C patients every 8 hours, escalating from 140 mg to 700 mg. A 5-fold increase in dose produced a 38-fold increase in silybin B exposure (Hawke 2010).
That is counterintuitive, but it means one thing: at customary doses the clearance pathway is nowhere near saturated, so almost everything is intercepted. Only by pushing far above customary doses does the pathway start to saturate and blood levels begin to hold.
So when you next see an in-vitro claim, ask one question first: is that concentration reachable by swallowing? The question applies to every polyphenol supplement — curcumin is the same story, see the curcumin island.
Depth · why phospholipid complexes exist
Given the wall, formulators have of course tried to go around it. You will see several claims on the shelf:Silybin-phospholipid complex. Complexing silybin with phosphatidylcholine helps it blend into the gut's lipid phase. Such formulations genuinely reach several-fold higher blood levels than plain extract — that part is measurable and real.
Added piperine, nanoparticles, liposomes. Same idea: raise absorption, or slow the phase II tagging.
But watch for a substitution here. Higher blood concentration is a pharmacokinetic endpoint; it proves more molecule got in. What you want to know is a clinical endpoint — did the liver get better. The first cannot answer for the second.
One counterexample settles it: the SyNCH trial used three to five times the customary dose, so blood levels were certainly not the limiting factor, and the result was still null (Fried 2012). Delivering more of it did not buy a better liver.
So the fair statement about enhanced formulations is: they solved the absorption problem, with no evidence they solved the efficacy problem. The premium buys the former.
Chapter 4
The IV version is a different drug
The IV version is a different drug
This is the most important scene on the island. The same molecule, sent in by a different route, becomes an antidote an emergency department genuinely uses — and what it does there is nothing like what the supplement ads describe.
Death cap toxin cannot enter a hepatocyte by itself
Alpha-amanitin, from Amanita phalloides, is extraordinarily toxic but has one weakness: it cannot cross the hepatocyte membrane. It has to borrow a door — on the blood-facing side of the hepatocyte sit two transporters, OATP1B3 and NTCP, whose day job is retrieving bile acids, and amanitin resembles a valid pass closely enough to be hauled in with them (Letschert 2006).
Once inside it locks RNA polymerase II, the pen the cell uses to transcribe its genes. With the pen pinned, the cell cannot even build proteins, and hepatocytes die in sheets.
Silibinin's job here is blocking the door, not antioxidation
It competitively occupies those two transporters so the toxin cannot get in. It also interrupts the toxin's enterohepatic circulation — excreted into bile, then reabsorbed from the gut, round and round.
Notice the mechanism changed. Antioxidation, glutathione restoration, quieting stellate cells — none of them is doing the work here. This is a transporter-competition story.
And the injection is not the same molecular form
The European intravenous product, Legalon SIL, contains silibinin dihemisuccinate — silibinin rebuilt as a water-soluble salt and injected directly. That step bypasses both walls from the previous scene: no gut to cross, and effective concentrations without first being tagged in the liver (Mengs 2012).
Be honest about this version's evidence too. Support comes largely from observational registry data, with no randomized controlled trial — nor should there be, since nobody will assign placebo to a mushroom-poisoning patient. The accurate statement is: clear mechanism, widespread clinical use, absent randomized evidence.
Take this judgment away: the shelf capsule and the emergency-room ampoule share a name, but they are a different molecular form, a different route, a different mechanism, and a different indication. The first cannot borrow the second's credibility. When an advertisement invokes mushroom poisoning to vouch for liver pills, those four differences are where you push back.
Death cap toxin cannot enter a hepatocyte by itself
Alpha-amanitin, from Amanita phalloides, is extraordinarily toxic but has one weakness: it cannot cross the hepatocyte membrane. It has to borrow a door — on the blood-facing side of the hepatocyte sit two transporters, OATP1B3 and NTCP, whose day job is retrieving bile acids, and amanitin resembles a valid pass closely enough to be hauled in with them (Letschert 2006).
Once inside it locks RNA polymerase II, the pen the cell uses to transcribe its genes. With the pen pinned, the cell cannot even build proteins, and hepatocytes die in sheets.
Silibinin's job here is blocking the door, not antioxidation
It competitively occupies those two transporters so the toxin cannot get in. It also interrupts the toxin's enterohepatic circulation — excreted into bile, then reabsorbed from the gut, round and round.
Notice the mechanism changed. Antioxidation, glutathione restoration, quieting stellate cells — none of them is doing the work here. This is a transporter-competition story.
And the injection is not the same molecular form
The European intravenous product, Legalon SIL, contains silibinin dihemisuccinate — silibinin rebuilt as a water-soluble salt and injected directly. That step bypasses both walls from the previous scene: no gut to cross, and effective concentrations without first being tagged in the liver (Mengs 2012).
Be honest about this version's evidence too. Support comes largely from observational registry data, with no randomized controlled trial — nor should there be, since nobody will assign placebo to a mushroom-poisoning patient. The accurate statement is: clear mechanism, widespread clinical use, absent randomized evidence.
Take this judgment away: the shelf capsule and the emergency-room ampoule share a name, but they are a different molecular form, a different route, a different mechanism, and a different indication. The first cannot borrow the second's credibility. When an advertisement invokes mushroom poisoning to vouch for liver pills, those four differences are where you push back.
Mechanism · why mushroom poisoning has a false recovery
The most dangerous feature of amatoxin poisoning is not the potency but the shape of its timeline — and that shape follows directly from the mechanism above.Why nothing happens for hours. The toxin pins the gene-transcribing pen; it does not shatter the cell on contact. Proteins already built keep working for a while, so for six to twenty-four hours after ingestion a person may feel nothing. By contrast, the mushrooms that make you violently ill immediately are usually far less lethal. Delayed onset is itself the warning sign.
Why there is a stretch that looks like recovery. The first wave is severe vomiting and diarrhoea, which then eases, and the person appears to have got through it. This is the false recovery period. In reality hepatocytes are dying in bulk, just not yet enough to show. Real liver failure surfaces a day or two later.
Why the antidote is a race. Blocking the door only helps while the toxin is still outside the cell. Once it is in and holding the polymerase, occupying doorways changes nothing. Every intervention is time-critical.
Red flag: vomiting and diarrhoea after eating wild mushrooms — whether or not there was an apparent recovery in between — means going to hospital immediately and telling the clinician that it was a wild mushroom and roughly when it was eaten. Bringing the leftovers or a photo helps. This is not wellness advice; it is an emergency.
Chapter 5
It has been tried in people
It has been tried in people
Mechanism done; now the human results. One thing to stress: this is not a case of insufficient evidence, so let's wait. The trials that needed doing were done, and they were negative.
Alcoholic and hepatitis B/C liver disease · Cochrane pooled
Cochrane pooled the randomized trials in these populations: whether combining all trials or restricting to the methodologically best ones, no significant effect on mortality or on complications of liver disease could be demonstrated (Rambaldi 2007).
Hepatitis C · the SyNCH trial
This one is worth remembering because it specifically closed an excuse. The previous scene showed that customary doses barely get in. So raise the dose. SyNCH used 420 mg or 700 mg three times daily — three to five times customary — for 24 weeks, double-blind and placebo-controlled.
Result: the decline in ALT did not differ across the three groups, and HCV RNA did not differ either (Fried 2012). The 'under-dosed' explanation is ruled out.
Steatohepatitis · the histology trial
The nonalcoholic steatohepatitis trial was run properly: 700 mg three times daily for a full 48 weeks, with the endpoint read from liver biopsy tissue, not from a blood panel. The primary endpoint — a 30% or greater fall in activity score without worsening fibrosis — was not met. There was a signal on fibrosis score, but that was a secondary endpoint (Wah Kheong 2017).
Why a secondary endpoint cannot be the conclusion. A trial measures many things at once, and the more it measures, the likelier one of them looks good by chance alone. The primary endpoint is the one nominated before starting as the decider; its value lies in being committed to in advance. Picking the pretty number out of the pile afterwards is marketing, not a conclusion. That judgment applies to every supplement claim.
Alcoholic and hepatitis B/C liver disease · Cochrane pooled
Cochrane pooled the randomized trials in these populations: whether combining all trials or restricting to the methodologically best ones, no significant effect on mortality or on complications of liver disease could be demonstrated (Rambaldi 2007).
Hepatitis C · the SyNCH trial
This one is worth remembering because it specifically closed an excuse. The previous scene showed that customary doses barely get in. So raise the dose. SyNCH used 420 mg or 700 mg three times daily — three to five times customary — for 24 weeks, double-blind and placebo-controlled.
Result: the decline in ALT did not differ across the three groups, and HCV RNA did not differ either (Fried 2012). The 'under-dosed' explanation is ruled out.
Steatohepatitis · the histology trial
The nonalcoholic steatohepatitis trial was run properly: 700 mg three times daily for a full 48 weeks, with the endpoint read from liver biopsy tissue, not from a blood panel. The primary endpoint — a 30% or greater fall in activity score without worsening fibrosis — was not met. There was a signal on fibrosis score, but that was a secondary endpoint (Wah Kheong 2017).
Why a secondary endpoint cannot be the conclusion. A trial measures many things at once, and the more it measures, the likelier one of them looks good by chance alone. The primary endpoint is the one nominated before starting as the decider; its value lies in being committed to in advance. Picking the pretty number out of the pile afterwards is marketing, not a conclusion. That judgment applies to every supplement claim.
Why the early small trials looked positive
Milk thistle accumulated a batch of positive small trials between the 1970s and 1990s, and that is the foundation of today's market. Three reasons explain why they looked positive without assuming anyone cheated.Transaminases fluctuate on their own. This is the biggest one, and the next scene is devoted to it. In chronic liver disease, ALT naturally rises and falls. If enrolment selects people whose values are currently high, then whatever you give them, the second draw tends to be lower than the first — regression to the mean. Without a control arm, that fall looks like drug effect.
Small samples wobble. With a few dozen participants, random variation dominates. Producing one attractive number is not hard.
Attractive results get published more often. Small negative trials frequently stay in a drawer, so the surviving literature is systematically skewed positive.
There is only one antidote to all three: large, randomized, double-blind, controlled, with the endpoint fixed in advance. That is exactly what the Cochrane pooling and SyNCH did, and their results were null.
This pattern recurs across supplements; the site's other textbook case is saw palmetto, where early small trials were uniformly positive and two large independent trials both failed. See the saw-palmetto island.
Chapter 6
A lower ALT is not a better liver
A lower ALT is not a better liver
Nearly every liver-pill pitch lands on the same sentence: my transaminases came down. That marker deserves its own scene, because it measures something other than what most people assume.
ALT leaks out; it is not secreted
ALT is an enzyme that lives inside hepatocytes, doing amino-acid metabolism. Normally it stays in the cell and only traces appear in blood.
So more of it in blood means something direct: hepatocytes broke, and the enzyme inside leaked out.
ALT reads leakage, not function. That single sentence yields three conclusions, each of which dismantles a class of claims.
One · it reflects whether damage is happening right now
It tells you injury is ongoing. It does not tell you how much working capacity the liver retains, nor how far fibrosis has progressed.
Two · in end-stage disease it can read normal
When most hepatocytes have already died and what remains is largely scar, there is nothing left to leak, and ALT can return to the normal range. That is the worst kind of normal. Watching this number alone means relaxing exactly when you should not.
Three · it wobbles by itself
Which connects to the previous scene: any intervention that involves drawing blood twice has a fair chance of showing a decline. Proving the intervention caused it requires a control arm.
What to look at instead
To know how the liver is actually doing, look at other things: synthetic capacity (albumin, clotting measures), clearance capacity (bilirubin), and degree of fibrosis (elastography or a fibrosis score). Those are outputs of the liver doing its job.
The hepatitis B island makes the same point from the other side: before the immune system engages, viral load can be enormous while ALT stays normal throughout — ALT measures how hard the immune system is fighting, not how much virus there is. See hepatitis-b.
ALT leaks out; it is not secreted
ALT is an enzyme that lives inside hepatocytes, doing amino-acid metabolism. Normally it stays in the cell and only traces appear in blood.
So more of it in blood means something direct: hepatocytes broke, and the enzyme inside leaked out.
ALT reads leakage, not function. That single sentence yields three conclusions, each of which dismantles a class of claims.
One · it reflects whether damage is happening right now
It tells you injury is ongoing. It does not tell you how much working capacity the liver retains, nor how far fibrosis has progressed.
Two · in end-stage disease it can read normal
When most hepatocytes have already died and what remains is largely scar, there is nothing left to leak, and ALT can return to the normal range. That is the worst kind of normal. Watching this number alone means relaxing exactly when you should not.
Three · it wobbles by itself
Which connects to the previous scene: any intervention that involves drawing blood twice has a fair chance of showing a decline. Proving the intervention caused it requires a control arm.
What to look at instead
To know how the liver is actually doing, look at other things: synthetic capacity (albumin, clotting measures), clearance capacity (bilirubin), and degree of fibrosis (elastography or a fibrosis score). Those are outputs of the liver doing its job.
The hepatitis B island makes the same point from the other side: before the immune system engages, viral load can be enormous while ALT stays normal throughout — ALT measures how hard the immune system is fighting, not how much virus there is. See hepatitis-b.
Clinical · the letters on the panel
This page is the lab-panel layer. How to interpret your own results must be judged by a clinician with your history; this only explains what each marker measures.ALT and AST · two enzymes that leak
ALT lives mainly in hepatocytes and is relatively liver-specific. AST also occurs in cardiac and skeletal muscle, so it rises after hard exercise too. Their ratio sometimes offers a clue, but a ratio is not a diagnosis.
ALP and GGT · the bile-duct route
These lean toward reflecting obstruction or irritation of the bile-export pathway. GGT is fairly sensitive to alcohol.
Albumin and clotting · these are the function ones
Albumin is synthesized by the liver, as are most clotting factors. When they fall, the liver's production line is in trouble, which is far more serious than a raised transaminase. Note that albumin's half-life is about three weeks, so it responds slowly and suits chronic assessment.
Bilirubin · clearance capacity
A breakdown product of red cells, processed by the liver and excreted into bile. A rise means that processing or export path is impaired; jaundice is it becoming visible in skin and eyes.
Elastography and fibrosis scores · how far the scarring has gone
These are what answer the fibrosis question, and fibrosis is what determines long-term outcome. Normal transaminases do not mean fibrosis has not progressed.
Red flags · when not to judge for yourself: yellowing of skin or eyes, markedly darker urine, persistent right upper abdominal pain, unusual bruising or gum bleeding, confusion or personality change. With any of these, stop all over-the-counter supplements and seek medical care promptly.
Chapter 7
Safe — but not the thing that works
Safe — but not the thing that works
After all that dismantling, the conclusion is not that it is harmful. Quite the opposite — among herbs its safety record is notably clean, and that has to be said plainly, because overstating risk misleads just as much as overstating benefit.
It barely harms the liver itself
The NIH drug-induced liver injury database rates milk thistle's likelihood of causing liver injury as category E — an unlikely cause of clinically apparent liver injury — and it has not been implicated in raising serum enzymes (LiverTox 2020). Few herbs earn that rating; Polygonum multiflorum and Senecio species are an entirely different matter, see the chinese-herb-safety island.
Common adverse effects
Orally it is well tolerated; the commonest complaints are digestive: bloating, nausea, gas, loose stools.
Take the allergy point seriously
It is an Asteraceae plant. People allergic to ragweed, chrysanthemum, marigold, or daisy may cross-react (NCCIH). With that history, self-experimenting is not advisable.
Drug interactions
In vitro it affects some hepatic metabolizing enzymes, but there is little evidence of clinically significant herb-drug interactions in people (LiverTox 2020). Even so, anyone on prescription medication — especially immunosuppressants, anticoagulants, or chemotherapy — should tell their clinician before adding any supplement.
So where is the real cost
Not in side effects. In substitution.
A bottle on the nightstand makes the liver feel like a handled problem. Meanwhile none of the things that actually change the outcome has been done: the drinking not stopped, the weight not lost, the hepatitis B never tested for, the hepatotoxic medication never reviewed.
What genuinely helps has its own islands here
How alcohol is dismantled in the liver, and why the first-step product is the damaging one, is on alcohol-metabolism. How fat accumulates in hepatocytes, and why weight loss is currently the only intervention with hard evidence, is on nafld. Why the virus cannot be fully cleared and what to monitor is on hepatitis-b. The liver's full clearance workshop, and where 'detox' stops being real, is on hepatic.
The judgment this island wants to leave you: facing any supplement claim, two questions suffice. First, at what concentration was this mechanism observed, and can swallowing reach it? Second, the marker being cited — does it measure function or leakage, and was it the primary endpoint or picked afterwards?
And as always: this page is education about mechanism, not medical advice. If you have liver disease, take medication, or have abnormal labs, take those questions to your doctor rather than to a shelf.
It barely harms the liver itself
The NIH drug-induced liver injury database rates milk thistle's likelihood of causing liver injury as category E — an unlikely cause of clinically apparent liver injury — and it has not been implicated in raising serum enzymes (LiverTox 2020). Few herbs earn that rating; Polygonum multiflorum and Senecio species are an entirely different matter, see the chinese-herb-safety island.
Common adverse effects
Orally it is well tolerated; the commonest complaints are digestive: bloating, nausea, gas, loose stools.
Take the allergy point seriously
It is an Asteraceae plant. People allergic to ragweed, chrysanthemum, marigold, or daisy may cross-react (NCCIH). With that history, self-experimenting is not advisable.
Drug interactions
In vitro it affects some hepatic metabolizing enzymes, but there is little evidence of clinically significant herb-drug interactions in people (LiverTox 2020). Even so, anyone on prescription medication — especially immunosuppressants, anticoagulants, or chemotherapy — should tell their clinician before adding any supplement.
So where is the real cost
Not in side effects. In substitution.
A bottle on the nightstand makes the liver feel like a handled problem. Meanwhile none of the things that actually change the outcome has been done: the drinking not stopped, the weight not lost, the hepatitis B never tested for, the hepatotoxic medication never reviewed.
What genuinely helps has its own islands here
How alcohol is dismantled in the liver, and why the first-step product is the damaging one, is on alcohol-metabolism. How fat accumulates in hepatocytes, and why weight loss is currently the only intervention with hard evidence, is on nafld. Why the virus cannot be fully cleared and what to monitor is on hepatitis-b. The liver's full clearance workshop, and where 'detox' stops being real, is on hepatic.
The judgment this island wants to leave you: facing any supplement claim, two questions suffice. First, at what concentration was this mechanism observed, and can swallowing reach it? Second, the marker being cited — does it measure function or leakage, and was it the primary endpoint or picked afterwards?
And as always: this page is education about mechanism, not medical advice. If you have liver disease, take medication, or have abnormal labs, take those questions to your doctor rather than to a shelf.
Shelf · what else is in a liver pill
Commercial liver products rarely contain milk thistle alone. Here are the usual companions, grouped by what they claim to do. This describes what these molecules do in the body and is not a recommendation to take anything; prescription products must be decided by a clinician.Also on the antioxidant route
Reduced glutathione and N-acetylcysteine appear frequently. Their oral predicament resembles silybin's but is not identical: swallowed glutathione is broken down to amino acids in the gut, whereas NAC supplies raw material for the body to build glutathione itself — a different strategy. NAC has a firmly established emergency role in acetaminophen overdose, which again is an intravenous, emergency, entirely different dose setting; do not conflate it with wellness dosing. See the nac and glutathione islands.
On the membrane-repair route
Polyenylphosphatidylcholine and similar products aim to supply phospholipid raw material to damaged hepatocyte membranes. Membrane composition is real biology; whether supplying it changes clinical outcome is a separate question, judged the same way as before — by the primary endpoint.
Prescription drugs sitting next to supplements
Some products contain genuine prescription ingredients. Their indication, dose, and duration are a clinician's call, not something to add at will.
Herbal formulas
This group needs the most caution, because natural does not mean safe. The site devotes a whole island to that sentence, including well-documented cases such as aristolochic-acid kidney injury and Polygonum multiflorum liver injury. See chinese-herb-safety.
One general heuristic: when a single product claims to protect the liver, lower enzymes, cure hangovers, fight fatigue, and boost immunity all at once, the breadth of the claim is itself the signal. Interventions that have actually been validated tend to have narrow indications.