Milk thistle

The least-weak liver signal. A meta-analysis shows modest ALT/AST reductions in NAFLD, and a 48-week biopsy trial in NASH hinted at a fibrosis benefit — but that same trial MISSED its primary histology endpoint. The effects are at the biomarker level (blood enzymes), not proven on hard outcomes.

Preliminary and surrogate-level. Lifestyle change and weight loss remain the only interventions with proven NAFLD/NASH outcome benefit; silymarin is at best adjunctive and unproven.

• Silymarin and liver enzymes in NAFLD (meta-analysis of 8 RCTs)

  positive · Meta-analysis

  Zhong et al., 2017, Medicine (Baltimore)n=587

  Pooling 8 RCTs (n=587) in NAFLD, silymarin modestly lowered ALT (−9.16 U/L) and AST (−6.57 U/L) versus control. A real but small biomarker-level effect — liver enzymes are a surrogate, not a hard outcome like fibrosis progression, cirrhosis, or mortality.

  Heterogeneous preparations and doses; effect sizes are small and at the level of blood enzymes, not clinical liver outcomes. Several included trials were small and unblinded.

  PubMed DOI

• Silymarin for biopsy-proven NASH (randomized, Malaysia)

  mixed · RCT

  Wah Kheong, Nik Mustapha & Mahadeva, 2017, Clinical Gastroenterology and Hepatologyn=99

  High-dose silymarin (2,100 mg/day) for 48 weeks did NOT meet its primary endpoint — it failed to reduce NAS by ≥30% more than placebo. However, a secondary biopsy finding showed more patients had ≥1-point fibrosis reduction on silymarin (22.4%) than placebo (6.0%). Encouraging on a secondary endpoint, but the trial missed its primary outcome.

  Primary endpoint was negative; the fibrosis result is a secondary outcome in a single ~99-patient trial and needs replication before being treated as established.

  PubMed DOI

Small RCTs and their meta-analysis suggest meaningful drops in fasting glucose (~27 mg/dL) and HbA1c (~1%) in type 2 diabetics on add-on silymarin. The point estimates are large, but the trials are small, heterogeneous, and the meta-analysis authors explicitly decline to make a recommendation.

Signal is in diagnosed type 2 diabetics on standard therapy, not healthy adults. Preliminary — not a substitute for prescribed glucose-lowering medication.

• Silymarin and glycemic control in type 2 diabetes (meta-analysis)

  positive · Meta-analysis

  Voroneanu et al., 2016, Journal of Diabetes Researchn=270

  Pooling 5 RCTs (n=270), silymarin reduced fasting glucose by ~27 mg/dL and HbA1c by ~1.0% in type 2 diabetics, with no effect on lipids. The point estimates look large, but the authors explicitly state the low quality and high heterogeneity of the trials mean no clinical recommendation can be made.

  The pooled trials are small, heterogeneous, and largely from one region; the authors themselves caution against drawing recommendations. Promising but preliminary.

  PubMed DOI

• Silymarin in type 2 diabetes (Huseini RCT)

  positive · RCT

  Huseini et al., 2006, Phytotherapy Researchn=51

  A 51-patient double-blind RCT added 600 mg/day silymarin to standard diabetes therapy. The silymarin group showed significant drops in HbA1c, fasting glucose, total cholesterol, LDL, triglycerides, and liver enzymes versus placebo. One of the larger contributors to the positive diabetes meta-analysis.

  Single small trial, add-on design; the strikingly broad benefit across many markers in a tiny sample warrants cautious interpretation pending larger replication.

  PubMed DOI

This is the headline claim and the weakest one. The Cochrane review found no convincing effect on mortality or histology in high-quality trials, and the NIH-funded SyNCH trial found oral silymarin no better than placebo for hepatitis C — even at doses far above supplement labels. For its single best-known use, the evidence is essentially null.

Do not use milk thistle as treatment for cirrhosis, alcoholic liver disease, or hepatitis B/C. The controlled evidence does not support it, and it is no substitute for medical care or alcohol cessation.

• Cochrane review — milk thistle for alcoholic and viral liver disease

  Null · Systematic review

  Rambaldi, Jacobs & Gluud, 2007, Cochrane Database of Systematic Reviewsn=915

  Cochrane pooled 13 RCTs (n=915) of milk thistle for alcoholic and/or hepatitis B/C liver disease. There was no convincing effect on all-cause mortality, liver-related mortality, complications, or histology. The conclusion: 'no evidence supporting or refuting' milk thistle — apparent benefits vanished when analysis was restricted to high-quality, adequately blinded trials.

  Most included trials were small and at high risk of bias; the few high-quality trials showed no effect. The famous 'milk thistle protects the liver' claim does not survive this review.

  PubMed DOI

• SyNCH — silymarin for hepatitis C (NIH-funded, JAMA)

  Null · RCT

  Fried et al., 2012, JAMAn=154

  The large, multicenter, NIH-funded, double-blind SyNCH trial tested higher-than-customary oral silymarin doses in 154 hepatitis C patients. Silymarin did not reduce ALT more than placebo, and there were no differences in HCV RNA or quality of life. Even at doses far above typical supplement labels, oral silymarin did nothing.

  Used a well-characterized standardized silymarin at supra-supplemental doses — this is close to a best-case test of the oral product, and it was flatly null.

  PubMed DOI