Herb–Drug Interactions (CYP450)

No human pharmacokinetic interaction study has been done. In rats, aloe leaf juice activated P-glycoprotein and CYP3A and lowered blood levels of cyclosporine. In laboratory tests the latex anthraquinone aloe-emodin is broken down by CYP3A4 and inhibits CYP1B1; neither has been confirmed in people.

Did not change blood levels of indinavir, a CYP3A4 substrate, or the glucuronidation and clearance of zidovudine in healthy volunteers. In a controlled trial it lowered warfarin blood levels and the INR, reducing warfarin's effect.

Leaf extracts inhibit CYP3A4 and CYP2C9 in laboratory tests and lowered CYP2C activity in rats, and andrographolide alone doubled warfarin exposure in rats. No human interaction study has been done.

A single dose of angelica root extract more than tripled caffeine exposure in six volunteers by inhibiting CYP1A2, an effect tied to its furanocoumarins. Effects on CYP3A4 and other enzymes have not been measured.

In laboratory tests, root and leaf extracts did not inhibit CYP1A2, CYP2C8, CYP2C9, CYP2D6 or CYP3A4, but some root extracts raised CYP3A4 expression in human liver cells. No human interaction study has been published.

In healthy volunteers, 28 days of ginseng lowered midazolam levels by a third (mild CYP3A4 induction); other studies saw no CYP3A4 change. Probe studies found no relevant effect on CYP1A2, CYP2C9, CYP2C19, CYP2D6, CYP2E1, P-glycoprotein or OATP1B1, and warfarin was unaffected.

In rats, oral astragalus granules roughly doubled liver CYP3A activity, and in mice the root and its main compounds raised liver P-glycoprotein. In human liver microsomes the extract inhibited CYP1A2 and CYP2C9. No human interaction study has been published.

Standardised extract inhibited CYP2C19, CYP2C9, CYP1A2 and CYP3A4 in laboratory tests, and in rats slowed CYP3A and P-glycoprotein, raising carbamazepine and digoxin levels. The isolated bacosides had little effect. No human study exists.

Bael fruit extract and its furanocoumarin marmelosin inhibited CYP3A4 (time-dependently) and CYP1A2 in laboratory tests with human liver enzymes; CYP2D6, CYP2C9 and CYP2C19 were unaffected. No study in animals or people has been done.

Berberine, barberry's main alkaloid, inhibited CYP3A4, CYP2D6 and CYP2C9 in healthy volunteers taking 900 mg a day, and raised ciclosporin levels in kidney-transplant patients; CYP1A2 and CYP2C19 were unaffected. Whole-barberry products have not been tested.

Extracts of five commercial uva-ursi products inhibited CYP3A4, other CYP3A enzymes and CYP2C19 in laboratory tests, and briefly inhibited, then induced, P-glycoprotein in cells. No study in animals or people has been done.

In healthy volunteers, two weeks of berberine (300 mg three times daily) inhibited CYP2D6, CYP2C9 and CYP3A4 but left CYP1A2 and CYP2C19 unchanged. It raised cyclosporine levels in kidney-transplant patients. Animal studies show it is a P-glycoprotein substrate.

Bergamottin, the peel and oil furanocoumarin, inhibits CYP3A4: 12 mg capsules raised felodipine levels about 40% in healthy volunteers. It also inactivates CYP2C9 in laboratory tests and inhibits CYP2D6 in mice. Furocoumarin-free oil lacks it; no study has tested bergamot itself.

Did not change CYP1A2, CYP2E1, CYP3A4 (midazolam) or P-glycoprotein (digoxin) activity in healthy-volunteer studies. A 7% dip in CYP2D6 in one study was not clinically relevant and was absent in another. CYP3A induction seen in mice did not involve the human receptor.

Piperine, its pungent alkaloid, inhibited CYP3A4, CYP2C9, CYP2E1 and P-glycoprotein in healthy volunteers taking 15–20 mg a day, raising levels of carbamazepine, diclofenac, fexofenadine, midazolam, phenytoin and theophylline. A curcumin–piperine product did not change paracetamol conjugation (UGT). Culinary pepper is untested.

In laboratory tests the root's alkaloid fraction strongly inhibited CYP1A2, CYP2C19, CYP2D6 and CYP3A4, while a whole methanol extract of the root had no effect. No interaction has been studied in people.

Its main alkaloid, sparteine, is cleared by CYP2D6, the enzyme behind the classic sparteine/debrisoquine poor-metaboliser trait, so poor metabolisers and people on CYP2D6 inhibitors may be more exposed. Broom itself has not been tested as an inhibitor.

Pulegone, a minor constituent of the leaf oil, is converted to the liver toxin menthofuran by CYP2E1, CYP1A2 and CYP2C19 in laboratory tests with human enzymes. No drug-interaction study of buchu itself has been done.

Saikosaponins lowered CYP3A4 expression and activity in human liver-derived cells, while saikosaponin d raised CYP1A2 and CYP2D6; in mice they induced CYP2E1 alongside liver injury. No pharmacokinetic study in people has been done.

No herb–drug interaction study of cocoa exists. Its theobromine and caffeine are cleared mainly through CYP1A2, so drugs that inhibit or induce CYP1A2, and smoking, change their clearance; pregnancy and oral contraceptives slow it.

Caffeine is cleared mainly by CYP1A2: fluvoxamine cut its clearance about fivefold in healthy volunteers, and smokers clear it faster than non-smokers.

An aqueous extract inhibited CYP3A by 99% at 375 µg/mL in a single laboratory test. It has not been studied in animals or people.

In laboratory tests the alcohol extract and its alkaloids inhibited CYP3A4, CYP2C9, CYP2C19 and CYP2D6, and activated PXR, raising CYP3A4 and CYP1A2 expression. The tea and its main alkaloid, californidine, had no effect. Not studied in people.

In healthy volunteers, a CBD-rich extract (640 mg CBD) inhibited CYP2C19, CYP2C9, CYP3A and CYP1A2, raising omeprazole, losartan, midazolam and caffeine levels, and repeated CBD tripled tacrolimus exposure. THC alone inhibited none of these, and is itself cleared mainly by CYP2C9.

No human interaction study of cascara has been done. In laboratory tests an infusion of the bark raised CYP3A4 gene expression in liver cells, apparently by activating the pregnane X receptor, and altered P-glycoprotein activity. Its anthraquinone emodin is oxidised in rat liver microsomes, partly by CYP1A2, to a more genotoxic metabolite. Neither finding has been confirmed in people.

No human interaction study has been published. In laboratory tests the bark's main compound, cinnamaldehyde, progressively inactivates CYP2A6, which clears coumarin, nicotine and letrozole, while CYP1A2, 2B6, 2C9, 2C19, 2D6 and 3A4 are far less affected; this has not been confirmed in people. A warfarin interaction is often assumed from cassia's coumarin, but coumarin itself is not an anticoagulant and no controlled study has shown it.

Inhibited CYP3A4, P-glycoprotein and OATP transporters in laboratory tests, though extracts also activated PXR and raised CYP3A4 expression in cells. In one patient, levels of atazanavir, ritonavir and saquinavir rose while taking it. No controlled human study exists.

In one rat study, catnip essential oil inhibited CYP2E1 and raised expression of the conjugating enzymes UGT and SULT, which protected the liver from paracetamol. It has not been studied in people.

Ceylon cinnamon (2 g three times daily) did not change nicotine or letrozole levels in healthy volunteers, despite predicted CYP2A6 inhibition. In laboratory tests cinnamaldehyde weakly inhibited CYP1A2 and CYP2C9, but not CYP3A4 or CYP2D6.

In laboratory tests a commercial chaste-tree berry preparation strongly inhibited CYP2C19 and CYP3A4, but not CYP1A1 or CYP2C9. No animal or human study has tested whether this changes drug levels.

In healthy volunteers the root extract slowed losartan metabolism (CYP2C9) and sped chlorzoxazone metabolism (CYP2E1), with no meaningful change in CYP3A4, CYP1A2, CYP2C19 or CYP2D6. Isolated baicalin induced CYP2B6 (bupropion) and lowered rosuvastatin levels; CYP3A inhibition is shown only in rats.

In rats CoQ10 sped up the clearance of warfarin, and in human liver microsomes it increased warfarin's 7-hydroxylation. In women with breast cancer, 300 mg a day left doxorubicin levels unchanged.

Caffeine is cleared mainly by CYP1A2. In healthy volunteers, fluvoxamine cut caffeine clearance about fivefold, ciprofloxacin doubled caffeine exposure, and oral contraceptives slowed its elimination. Heavy coffee drinking is linked to higher CYP1A2 activity in some people.

In mice, a standardised root extract induced liver CYP3A, CYP2B and CYP2C enzymes, mainly through constituents other than forskolin; the effect reversed soon after stopping. Forskolin raised CYP3A4 in mice carrying the human gene. No human study has been published.

Berberine, its main alkaloid, slowed CYP2D6, CYP2C9 and CYP3A4 in healthy volunteers and raised ciclosporin levels in kidney-transplant patients. A whole coptis decoction did the opposite in rats, lowering ciclosporin by activating P-glycoprotein and CYP3A.

Curcumin 1 g daily cut CYP1A2 activity by about 29% in healthy men and raised sulfasalazine (BCRP) levels; a curcumin–piperine extract left midazolam (CYP3A4) and flurbiprofen (CYP2C9) unchanged. Talinolol (P-glycoprotein) results conflict. Curcumin itself is cleared by glucuronidation.

In healthy volunteers a single dose briefly raised midazolam levels, but 10–14 days of danshen extract induced CYP3A4 and intestinal P-glycoprotein, lowering midazolam (by up to 80%), fexofenadine and clopidogrel. Theophylline (CYP1A2) was unchanged. Danshen strengthens warfarin, with bleeding cases on record.

In rats, four weeks of dandelion tea lowered liver CYP1A2 and CYP2E1 activity and more than doubled UGT activity, with no change in CYP2D or CYP3A. No human study has tested it.

Some preparations weakly inhibited CYP3A4 in laboratory tests, at concentrations the authors judged clinically unlikely to matter, and both blocked and upregulated P-glycoprotein in kidney cells. Harpagoside itself had little effect. No animal or human interaction study exists.

Prescription omega-3 (EPA plus DHA, 4 g a day) did not change simvastatin or atorvastatin levels, both cleared by CYP3A4, in healthy volunteers; liver-microsome tests predict no clinically meaningful CYP450 inhibition.

An extract inhibited CYP2C19 in human liver microsomes, and in rats a water extract raised CYP3A and CYP2D activity. No human study exists. Its warfarin interaction appears pharmacodynamic: in rabbits it changed clotting without changing warfarin levels.

Induced CYP3A in healthy volunteers, lowering midazolam levels by about a quarter, and modestly inhibited CYP1A2 (caffeine) in one study. It did not meaningfully change CYP2C9, CYP2D6, CYP2E1 or P-glycoprotein, or levels of lopinavir–ritonavir, darunavir–ritonavir, etravirine or docetaxel.

Did not change CYP2D6 (dextromethorphan) or CYP3A4 (alprazolam) activity in 12 healthy volunteers taking a standardised extract for 14 days. Separately, eleuthero can distort some digoxin blood tests.

5-MeO-DMT is cleared mainly by MAO-A; CYP2D6 converts a smaller share to bufotenine, shown in human liver tests and mice. The Virola lignan grandisin inhibited CYP2C9 and CYP3A4 in laboratory tests.

In rats, ephedra water decoction induced CYP1A2 and lowered blood levels of theophylline. In people, ephedrine is mostly excreted unchanged in urine. No human CYP study exists; its important interactions (MAO inhibitors, caffeine) act on the body, not on metabolism.

A water extract inhibited CYP3A4, CYP1A2 and CYP2D6, partly time-dependently, in laboratory tests, and the root rotenoid boeravinone B weakly blocks P-glycoprotein in cells. No animal or human interaction study exists.

1,8-Cineole, the main constituent of the oil, is broken down mainly by CYP3A4 in human liver tests, so drugs that block or induce CYP3A4 could change its levels. No study has tested eucalyptus with other drugs.

Fennel extract inhibited CYP2D6 and CYP3A4 in laboratory tests, the latter partly through bergapten (5-methoxypsoralen). In rats it lowered ciprofloxacin levels, probably by binding the drug to minerals in the extract rather than through enzymes. No human study exists.

A week of fenugreek seed did not change CYP2D6 or CYP3A4 activity in healthy volunteers, and it left ciclosporin and carbamazepine levels unchanged in rabbits. Laboratory inhibition was weak. Single case reports involve warfarin (with boldo) and ribociclib.

At 4 g a day, prescription omega-3 ethyl esters did not change simvastatin or atorvastatin levels (CYP3A4), and pure EPA did not change omeprazole, rosiglitazone or warfarin; the label expects no clinically significant CYP450 inhibition.

Digitoxin, the main foxglove glycoside, is cleared by CYP3A in human liver tissue and carried by P-glycoprotein in laboratory tests. Rifampicin, which induces both, sharply lowered digitoxin levels in patient reports.

Resin extracts of four Boswellia species inhibited CYP1A2, CYP2C8, CYP2C9, CYP2C19, CYP2D6 and CYP3A4 in laboratory tests, and the extract and its keto-boswellic acids inhibited P-glycoprotein in cells. No interaction study in people has been done.

Garlic oil lowered CYP2E1 activity by 22–39% in healthy volunteers, while garlic products left CYP1A2, CYP2D6 and CYP3A4 unchanged. Garlic supplements halved saquinavir levels in volunteers, probably by raising intestinal P-glycoprotein.

Gentiopicroside, the main bitter compound, inhibited CYP2A6 but not CYP1A2, CYP2C9, CYP2D6 or CYP3A4 in laboratory tests, and induced none in human liver cells. Amarogentin raised CYP3A4 expression in liver cells. The EU monograph reports no interactions.

Chamomile essential oil and its constituents (chamazulene, spiroethers, α-bisabolol) inhibited CYP1A2 most strongly, and CYP3A4, CYP2C9 and CYP2D6 more weakly, in laboratory tests. Chamomile tea lowered CYP1A2 activity in rats. No human pharmacokinetic study exists.

Did not change warfarin levels, CYP2C9 activity or INR in healthy volunteers. In laboratory tests ginger extract and its gingerols and shogaols inhibited CYP3A4, CYP1A2, CYP2B6 and P-glycoprotein; ginger juice raised tacrolimus levels in rats. Neither has been confirmed in people.

Healthy-volunteer probe studies found no relevant effect on CYP1A2, CYP2B6, CYP2C9, CYP2D6, CYP2E1 or CYP3A4, and no change in warfarin or digoxin levels. Single studies saw repeated dosing modestly raise talinolol (P-glycoprotein) and lower omeprazole (CYP2C19) levels.

Inhibits CYP3A and CYP2D6 in healthy volunteers, raising midazolam levels 40–60% and roughly halving CYP2D6 activity. It lowered metformin absorption but did not change digoxin (P-glycoprotein), rosuvastatin (OATP) or indinavir levels. CYP2C8 and CYP2C9 inhibition is laboratory-only.

In laboratory tests gotu kola extracts weakly inhibited CYP2C9 and CYP2C19, with little effect on CYP2D6 or CYP3A4, and a water extract did not induce CYP1A2, CYP2B6 or CYP3A4. No interaction has been studied in people.

Inhibited CYP3A4 and CYP2C9 in laboratory tests, with large differences between brands. Three months of a grape seed procyanidin extract did not change CYP3A4 activity, measured by cortisol metabolism, in heavy and former smokers.

Guarana's caffeine is cleared mainly by CYP1A2, so CYP1A2 inhibitors such as fluvoxamine and ciprofloxacin raise caffeine levels in people. Guarana's own effect on drug-metabolising enzymes has not been studied in humans.

Its caffeine is cleared mainly by CYP1A2, so CYP1A2 inhibitors such as fluvoxamine, which cut caffeine clearance about fivefold in volunteers, can make the drink much stronger and longer-lasting. Guayusa itself has not been tested.

No human interaction studies. In rats, a week of ethanolic leaf extract raised blood levels of phenacetin (CYP1A2) about 1.4-fold and lowered those of tolbutamide (CYP2C9), with no effect on amlodipine (CYP3A4).

In laboratory tests a gypenoside extract inhibited CYP2D6 strongly, and CYP2C8, CYP3A4 and CYP2C9 more weakly. No animal or human study has tested whether this matters in practice.

In healthy volunteers, three weeks of hawthorn extract WS 1442 did not change blood levels of digoxin, a P-glycoprotein substrate. No CYP enzyme study has been done, and the EU monograph reports no interactions.

No pharmacokinetic interaction study has been done in people. In human liver microsomes, tulsi's main leaf-oil constituent eugenol inhibited CYP1A2, CYP2C9, CYP2D6 and CYP3A4, and in goats raw leaf juice raised the bioavailability of intravenous ceftriaxone. Neither has been confirmed with the whole herb in people.

A standardised hop extract taken twice daily for two weeks did not change CYP1A2, CYP2C9, CYP2D6 or CYP3A4 activity in a clinically relevant way in 16 peri- and postmenopausal women; alprazolam exposure fell 7.6%. Laboratory reports of inhibition did not translate to people.

In rats given aescin by injection, CYP1A2 activity rose while CYP2C9 and CYP3A4 activity fell. Oral horse chestnut has not been tested, no human study exists, and the EU monograph reports no interactions.

Increased CYP3A4 gene expression in liver cells in laboratory tests, with no effect on P-glycoprotein. Two case reports describe HIV treatment failure on efavirenz-based regimens after a horsetail supplement was added; no pharmacokinetic study has tested this.

Huperzine A is cleared largely unchanged by the kidneys: human liver microsomes did not break it down, and it barely inhibited seven CYPs at clinical levels. One cell study found CYP3A4 induction; cell assays show it is a P-gp substrate.

The oil's convulsant ketones, cis- and trans-3-pinanone, are broken down by CYP3A4 in human liver microsomes, a step that makes them less toxic. No interaction study has been done in animals or people.

A juniper-berry decoction inhibited CYP3A4 in a laboratory test, almost completely after 60 minutes of boiling. No animal or human study has tested it, and the EU herbal monograph carries no interaction warning.

Lowered CYP2E1 activity by about 40% in healthy volunteers but did not change CYP2D6, CYP3A4 (midazolam) or P-glycoprotein (digoxin). Laboratory tests also show inhibition of CYP2B6, CYP2C9 and CYP2C19.

A single 2 g dose of kratom tea raised midazolam exposure about 40% in healthy volunteers, mainly by inhibiting CYP3A in the gut, and did not change CYP2D6. Mitragynine is itself converted by CYP3A4 to 7-hydroxymitragynine; itraconazole slows this.

Silexan lavender oil (160 mg/day) did not change CYP1A2, CYP2C9, CYP2C19, CYP2D6 or CYP3A4 activity in a healthy-volunteer probe study, nor levels of an ethinylestradiol/levonorgestrel pill. Weak CYP3A4 and CYP1A2 inhibition in laboratory tests did not translate to people.

Glycyrrhizin (300 mg/day for 14 days) modestly induced CYP3A4 in healthy volunteers, lowering midazolam and omeprazole levels, but did not change P-glycoprotein or CYP2C19. A glycyrrhizin-depleted extract left CYP1A2, CYP2C9, CYP2D6 and CYP3A4 unchanged in women.

Lobeline, the main alkaloid, inhibited P-glycoprotein drug efflux in drug-resistant tumour cells in laboratory tests. The effect of the whole herb has not been tested, and no CYP enzyme or human interaction study of lobelia has been found.

Tobacco smoke induces CYP1A2 (and CYP2E1) in people through its combustion products, speeding clearance of theophylline, caffeine, imipramine and haloperidol; shown for cigarette tobacco, and snuff or chewed tobacco does not do it. Nicotine itself is cleared by CYP2A6 and UGT.

Its alkaloid colchicine is cleared by CYP3A4 and pumped out by P-glycoprotein. In people, clarithromycin, ketoconazole, ritonavir, ciclosporin, verapamil and diltiazem raised colchicine levels enough to need 33–75% dose cuts; grapefruit juice raised its absorption in cell and rat tests.

Melatonin is cleared mainly by CYP1A2: fluvoxamine raised its exposure about 17-fold and caffeine about twofold in healthy volunteers, and smoking lowers it. CYP2C19 and CYP2C9 contribute in lab studies; the licensed product label notes CYP3A induction only at supra-therapeutic lab concentrations.

Usual silymarin doses did not change CYP1A2, CYP2D6, CYP2E1 or CYP3A4 activity in healthy volunteers, nor irinotecan, digoxin or rosuvastatin levels. Silymarin modestly raised losartan (CYP2C9) and talinolol (P-glycoprotein) levels in two small studies.

Modafinil induces CYP3A4: in volunteers it cut triazolam exposure by 59% and lowered ethinyl estradiol, so hormonal contraceptives can fail. It inhibits CYP2C19, raising drugs such as omeprazole, phenytoin and diazepam. CYP3A4 also helps clear it.

In laboratory tests on human liver cells, myrrh resin extracts raised expression of CYP2C9, CYP2C8, CYP2C19 and CYP3A4, apparently by activating the pregnane X receptor. No animal or human study has tested whether this lowers drug levels.

In human liver microsomes nefiracetam is hydroxylated to its main metabolite chiefly by CYP3A4, with a minor CYP1A2 share. No interaction study in people has been published.

Black seed (2.5 g twice daily for a week) slowed CYP2D6 and CYP3A4 breakdown of dextromethorphan in a four-person volunteer study. Thymoquinone inhibits CYP2C9 most potently in laboratory tests. In animals, black seed lowered ciclosporin and phenytoin levels.

In laboratory tests, myristicin is a mechanism-based inhibitor of CYP1A2 and is itself broken down by CYP3A4 and CYP1A2; nutmeg and mace extracts inhibited CYP3A4 or CYP2C9. In rats, injected myristicin induced several liver CYPs. No human study.

A passionflower infusion raised CYP3A4 gene expression in liver cells in a laboratory test, without activating PXR or affecting P-glycoprotein. No animal or human study has measured an effect on drug levels.

Phenylpropanoid glycosides isolated from the bark moderately inhibited CYP3A4 in laboratory tests. A bark infusion did not change CYP3A4 gene expression or P-glycoprotein activity in cell tests. No study in animals or people has been done.

Paeoniflorin, the root's main glycoside, moderately inhibited CYP3A in rats, though a second rat study found it induced CYP1A2. In laboratory tests 4-O-galloylalbiflorin inhibited CYP3A4, CYP2C9 and CYP2D6. No human interaction study exists.

Peppermint oil (600 mg) raised felodipine levels by about 40% in healthy volunteers, consistent with CYP3A4 inhibition; menthol alone did not. In mice menthol increased CYP2C and CYP3A activity and lowered warfarin levels.

In human liver microsomes pterostilbene strongly inhibits CYP2C9 and also inhibits CYP2C8, and it is broken down mainly by CYP1A2. It inhibits several UGT enzymes, most potently UGT1A9. No interaction study in people has been published.

Raspberry leaf extract inhibited CYP1A2, CYP2D6 and CYP3A4 in laboratory tests with human enzymes, at levels the authors judged possibly relevant at usual doses. No interaction has been studied or reported in people.

Did not change blood levels of caffeine, tolbutamide, dextromethorphan or alprazolam — test drugs for CYP1A2, CYP2C9, CYP2D6 and CYP3A4 — in 15 women taking 120 mg of isoflavones a day for two weeks.

Its monacolin K is lovastatin, cleared by CYP3A4: itraconazole and grapefruit juice raise lovastatin levels many-fold in people, and gemfibrozil raised lovastatin acid taken as red yeast rice. One capsule did not alter nifedipine levels.

Its iridoid catalpol inhibited CYP3A4, CYP2C9 and CYP2E1 in human liver microsomes and left five other enzymes unaffected. This has not been tested with the whole root in people; in rats, rehmannia extract did not change clozapine levels.

Not tested in people. In rat liver microsomes a reishi polysaccharide inhibited CYP2E1, CYP1A2 and CYP3A, and in drug-resistant leukaemia cells it lowered expression of the drug pump P-glycoprotein. Neither finding has been confirmed in humans.

Fourteen days of rhodiola lowered CYP2C9 activity by about 21% in 13 healthy volunteers, relevant to warfarin and phenytoin. The same study found no change in CYP1A2, CYP2C19, CYP2D6 or CYP3A4, despite strong CYP3A4 and P-glycoprotein inhibition in laboratory tests.

In rats, two weeks of rooibos tea cut blood levels of midazolam by more than 60% by inducing intestinal CYP3A. In laboratory tests, extracts instead inhibit CYP3A4, CYP2C8, CYP2C9 and CYP2C19. No human interaction study has been published.

In laboratory tests rosemary extract blocked P-glycoprotein, and its diterpene carnosic acid both inhibited CYP2C9 and CYP3A4 and induced CYP3A4 and CYP2B6 in human liver cells. Rosemary extracts induced UGT in rats. No human interaction study exists.

A six-day saffron infusion slightly lowered CYP1A2 activity in healthy volunteers, significantly only in men, and did not change CYP2A6. Laboratory and animal studies report effects on CYP3A4, CYP2E1 and CYP1A, with CYP3A4 inhibition the most consistent, but these have not been confirmed in people. A week of 200–400 mg a day did not change clotting tests.

Thujone, from the leaf oil, is cleared mainly by CYP2A6, then CYP3A4 and CYP2B6, in human liver preparations. Two weeks of sage tea raised liver CYP2E1 in mice. No human drug-interaction study has been done.

Only animal and lab data exist: in rats, a week of oral SAMe sped up a CYP1A2 probe drug and slowed CYP2D6, CYP3A4, CYP2C19 and CYP2C9 probes, and in microsomes it weakly inhibited CYP2E1. No human study has tested it.

In laboratory tests, astilbin and its isomers (flavonoids of Smilax glabra) inhibited CYP3A4 and CYP2D6, and a S. china root extract induced the nicotine-clearing enzyme CYP2A6 in liver cells. No animal or human interaction study exists.

Did not change CYP1A2, CYP2D6, CYP2E1 or CYP3A4 activity in two healthy-volunteer probe studies, including alprazolam and dextromethorphan tests. Interactions through these enzymes look unlikely at usual doses.

Its close relative Schisandra sphenanthera inhibits CYP3A: its extracts roughly doubled midazolam and tacrolimus levels in volunteers and transplant patients. S. chinensis extract raised talinolol levels 47% in volunteers (P-glycoprotein). The lignan macelignan inactivates CYP2C19 in rats.

In rats, root extract lowered liver CYP2E1 activity while protecting against isoniazid liver injury, and a water extract did not inhibit CYP3A4 in rat liver microsomes. No human interaction study exists.

Induces CYP3A4, CYP2C9, CYP2C19, CYP2B6, CYP2E1 and P-glycoprotein in people, through hyperforin, lowering levels of ciclosporin, warfarin, digoxin, omeprazole, oral contraceptives and many other drugs. It does not affect CYP2D6 or CYP1A2. Extracts with no more than 1 mg hyperforin a day have not shown the effect.

Harmine and harmaline inhibit CYP2D6 and CYP3A4 in laboratory tests and are themselves cleared mainly by CYP2D6 and CYP1A2; CYP2D6 status changed harmaline exposure in mice. Seed extract raised CYP1A2, CYP2C19 and CYP3A4 gene expression in liver cells. MAO-A inhibition remains the dominant interaction.

Two alkamides isolated from the leaf are mechanism-based inhibitors of CYP3A4 in laboratory tests (IC50 3–10 µM). Whether tarragon as eaten or taken affects CYP3A4 drugs has not been tested.

In diabetic rats and acetaminophen-poisoned mice, taurine lowered CYP2E1 activity and expression. In liver cells it doubled rifampicin's induction of CYP3A4 without inducing it alone. No human study has tested it on any CYP enzyme.

Green tea cut nadolol levels by 85% in healthy volunteers, likely by blocking the gut transporter OATP1A2, and modestly lowered atorvastatin and digoxin. Catechin extracts left CYP1A2, CYP2C9, CYP2D6 and CYP3A4 unchanged. Tea's caffeine is cleared by CYP1A2, which fluvoxamine blocks.

In healthy volunteers tesamorelin barely changed levels of simvastatin, a CYP3A substrate (AUC down 8%), or ritonavir (down 9%). Its label notes that the growth hormone it releases may alter clearance of CYP450-metabolised drugs such as corticosteroids, sex steroids, anticonvulsants and cyclosporine.

Thyme inhibited CYP3A4 and several other drug-metabolising enzymes in laboratory tests, and raised enzyme gene expression in rats. Thymol itself is cleared mainly by CYP2A6. No interaction has been seen in people.

Repeated doses of Desmodium adscendens leaf extract shortened zoxazolamine paralysis and blocked thiopentone-induced sleep in rats, signs of induced liver CYP enzymes; no specific enzyme was identified. Not studied in people.

A methanol extract blocked P-glycoprotein drug efflux in drug-resistant breast-cancer cells in laboratory tests. No animal or human interaction study exists, and no CYP data were found.

Smoking induces CYP1A2 (and CYP2E1) through combustion products, not nicotine, speeding clearance of theophylline, caffeine, imipramine, haloperidol and others; snuff and chewed tobacco do not. Nicotine itself is cleared by CYP2A6 and UGT.

Curcumin inhibited CYP1A2 in healthy volunteers and raised sulfasalazine levels threefold through the BCRP transporter. Curcumin products did not change midazolam or nifedipine (CYP3A4), flurbiprofen (CYP2C9) or paracetamol conjugation in people, despite CYP3A4 inhibition in laboratory tests. Talinolol (P-glycoprotein) results conflict.

Did not change CYP1A2, CYP2D6, CYP2E1 or CYP3A4 activity in healthy-volunteer probe studies; 1,000 mg nightly raised alprazolam's peak level slightly without changing overall exposure. A rat study showing intestinal CYP3A inhibition has not been borne out in people.

Vinpocetine strongly inhibits P-gp in cell tests, while its CYP3A4 and CYP2D6 inhibition needs levels far above those in blood. In people it left imipramine, oxazepam and glibenclamide levels unchanged and barely altered warfarin's effect.

In three epilepsy patients, gram doses of nicotinamide slowed primidone's conversion to phenobarbital and raised carbamazepine levels. In lab tests, nicotinic acid and nicotinamide weakly inhibit CYP2D6, nicotinamide also CYP3A4 and CYP2E1, and CYP2E1 oxidises nicotinamide.

Vitamin C at 1 g a day did not change the clearance of caffeine, a CYP1A2 marker, or of antipyrine in elderly men. The same dose lowered indinavir levels about 14% in healthy volunteers, by an unestablished mechanism.

Vitamin D2 is activated in the liver mainly by CYP2R1, an enzyme outside the drug-metabolising set. In lab studies CYP3A4 also hydroxylates it, though weakly.

Vitamin D3's active forms are broken down partly by CYP3A4, which rifampin speeds up. In people, 800 IU daily lowered atorvastatin levels and repletion raised midazolam clearance, though two trials using a CYP3A4 blood marker found no change.

Vitamin E is broken down by CYP4F2, shown with human liver microsomes and recombinant enzymes. Very high injected doses doubled CYP3A and P-gp in rat liver, and tocotrienols switched on CYP3A4 in liver cells.

Its saponins diosgenin and dioscin inhibited CYP3A4, but not CYP1A2, CYP2C9 or CYP2D6, in laboratory tests, and diosgenin is pumped out of gut cells by P-glycoprotein. No interaction has been studied in people.

Willow bark extracts inhibited CYP1A2, CYP3A4 and CYP2C19 in human liver cells only at high concentrations, and white willow infusion raised CYP3A4 gene expression in another cell study. No human study exists; the warfarin caution in the EU monograph is not a CYP effect.

Thujone is broken down mainly by CYP2A6 in human liver preparations, with minor roles for CYP3A4 and CYP2B6, and it inhibits CYP2A6 and CYP2B6 in laboratory tests. No human drug-interaction study has been done.

In a laboratory study a standardised maté extract inhibited CYP3A4; this has not been tested in people. Its caffeine is cleared mainly by CYP1A2, so medicines that block CYP1A2 can raise caffeine levels.

Spinosin and the other seed flavonoids are pumped back out of intestinal cells by P-glycoprotein in laboratory tests, which limits their absorption. No CYP450 or human interaction study exists.