Ren Shen Ye

Randomised, double-blind, placebo-controlled trial in 20 healthy volunteers given warfarin for three days in weeks 1 and 4, with ginseng or placebo from week 2. Peak INR fell significantly with ginseng relative to placebo (difference -0.19, 95% CI -0.36 to -0.07, P = 0.0012). Important caveat for this record: the herb was American ginseng, Panax quinquefolius root, not Panax ginseng, and not the leaf.

Randomised, open-label, controlled study in 25 warfarin-naive patients with newly diagnosed ischaemic stroke, allocated to warfarin plus Panax ginseng (n = 12) or warfarin alone (n = 13) for two weeks. INR and prothrombin time rose from baseline in both groups as expected, but there was no statistically significant difference in peak values or AUC between groups in either week. The authors conclude that co-administration does not influence warfarin's pharmacologic action. This is the trial that actually used Panax ginseng root; it points the opposite way to the American ginseng study.

Cochrane review of nine randomised double-blind placebo-controlled trials, of which only five had extractable data, all in healthy participants. Pooling was impossible because of heterogeneity in outcomes, duration and dose. Some aspects of cognition, behaviour and quality of life appeared to improve and no serious adverse events were attributed to ginseng, but the reviewers concluded there is a lack of convincing evidence for a cognitive-enhancing effect in healthy people and no high-quality evidence in dementia. All included preparations were root-based.

Seven randomised trials of Korean red ginseng in erectile dysfunction, of low average methodological quality. Meta-analysis of the six placebo-controlled trials favoured red ginseng (n = 349, risk ratio 2.40, 95% CI 1.65 to 3.51, P less than 0.00001, I-squared 22%), with a similar effect in the psychogenic subgroup. The authors call the evidence suggestive rather than definitive given the small pooled sample and poor trial quality. Red ginseng is steamed root, not leaf.

Twelve healthy volunteers aged 60 to 76 took St John's wort, garlic oil, Panax ginseng and Ginkgo biloba for 28 days each with 30-day washouts, with midazolam, caffeine, chlorzoxazone and debrisoquine probes before and after. Panax ginseng inhibited CYP2D6 significantly but by only about 7%, which the authors judged not clinically relevant, and had no effect on CYP1A2, CYP2E1 or CYP3A4. St John's wort, by contrast, induced CYP3A4 by around 140%.

Open-label, randomised, three-period study in 15 healthy men given a probe cocktail (caffeine, losartan, omeprazole, dextromethorphan, midazolam, pitavastatin) with single or 15-day dosing of red ginseng extract. Probe pharmacokinetics were comparable to control, with AUC ratios inside 0.8 to 1.25 for the CYP2C9, CYP3A4 and OATP1B1 probes and only slightly higher for CYP1A2, CYP2C19 and CYP2D6 probes. Parallel microsome, hepatocyte and transporter assays on the extract and on ginsenoside Rb1 were negative. Conclusion: minimal risk of clinically relevant CYP- or OATP-mediated interaction from red ginseng root.

One of the few toxicology studies on the leaf drug rather than the root. A powdered ginseng leaf extract was negative in the Ames test across four Salmonella typhimurium strains and one Escherichia coli strain with and without metabolic activation, produced no increase in chromosomal aberrations, and at oral doses up to 2000 mg/kg in male mice did not raise micronucleated polychromatic erythrocytes or cause clinical signs, weight loss or mortality. The authors note explicitly that toxicity information on ginseng leaf is otherwise insufficient for clinical use.

Leaf-specific pharmacology in a high-fat-diet rat model, reporting reduced weight gain and improved metabolic parameters with Panax ginseng leaf extract. Cited here as the basis for the theoretical additive effect with glucose-lowering drugs; it is an animal study and does not support a clinical metabolic indication.

Analysed the ginsenoside content of American ginseng leaf and tested its hypoglycaemic activity in animals, finding a leaf ginsenoside profile distinct from the root and a glucose-lowering effect. Included because it is the clearest demonstration that a Panax leaf is a chemically and pharmacologically different drug from its root; the species is Panax quinquefolius, not Panax ginseng, so it does not transfer directly.

Systematic appraisal of the safety data reported in randomised clinical trials of Panax ginseng. The reviewers emphasise that trial results conflict and depend heavily on protocol design, participant characteristics and which type of ginseng was used, red, white, fermented or cultured. The trials it draws on are of root preparations; there is no comparable safety review of the leaf drug.