Substance + claimed effects

This entry covers two common dietary substances — grapefruit (and related citrus: pomelo, Seville orange, tangelo) and St. John's wort (Hypericum perforatum) — viewed through a single pharmacological lens: their opposite-direction modulation of cytochrome P450 3A4 (CYP3A4), the liver and gut enzyme responsible for first-pass metabolism of roughly half of all prescription drugs. Grapefruit's furanocoumarins (notably bergamottin and 6',7'-dihydroxybergamottin) irreversibly inhibit intestinal CYP3A4, raising plasma concentrations of affected drugs; St. John's wort's hyperforin activates the pregnane X receptor and induces hepatic and intestinal CYP3A4 (and the efflux transporter P-glycoprotein), lowering plasma concentrations Bailey 2013 Moore et al. 2000. The drug classes the description names — statins, immunosuppressants, calcium-channel blockers, hormonal contraceptives, psychotropics — share heavy CYP3A4 dependence and narrow therapeutic windows. Consequences covered here are iatrogenic harm and therapeutic failure: rhabdomyolysis from raised statin levels, acute transplant rejection from suppressed tacrolimus or cyclosporine levels, contraceptive failure with unintended pregnancy, supratherapeutic calcium-channel blocker effects (hypotension, bradycardia), and loss of antiretroviral / antidepressant / anxiolytic efficacy. The article scores against the avoidance behaviour (don't eat grapefruit / don't take SJW while on these drugs), not against the drugs themselves.

Evidence by addressing question

Mechanism

Grapefruit — irreversible inhibition of intestinal CYP3A4. The original discovery was accidental: Bailey et al. 1991 used grapefruit juice to mask the taste of ethanol in a felodipine pharmacokinetic study and found felodipine area-under-curve (AUC) tripled. The mediating compounds are furanocoumarins — bergamottin, 6',7'-dihydroxybergamottin, and bergaptol — which form covalent adducts with CYP3A4 in the enterocytes of the small intestine. Lown et al. 1997 showed that a single 8-oz glass of grapefruit juice reduced intestinal CYP3A4 protein content by ~62% on duodenal biopsy, with no change in hepatic CYP3A4 mRNA. Definitive causality was established with a furanocoumarin-free grapefruit juice trial: stripping the furanocoumarins eliminated the felodipine interaction entirely, confirming furanocoumarins as the mediators rather than flavonoids (naringin, naringenin) — a long-standing misattribution. Inhibition is mechanism-based / suicide: the enzyme is destroyed, so recovery requires de novo synthesis. Greenblatt et al. 2003 quantified recovery: ~50% of CYP3A4 activity returns within 24 h after a single dose, but full recovery takes ~72 h; a hypothesis sometimes called the "three-day rule."

St. John's wort — induction via the pregnane X receptor (PXR / NR1I2). Moore et al. 2000 demonstrated that hyperforin is one of the most potent known activators of the human PXR (EC50 ~23 nM), a nuclear receptor whose transcriptional targets include CYP3A4 (~700% induction in primary hepatocytes), CYP2C9, CYP1A2, and the efflux transporter MDR1/P-glycoprotein. Induction is slow on, slow off: enzyme upregulation requires sustained dosing (typically 10–14 days of 900 mg/day standardised extract to reach peak induction) and decays over 1–2 weeks after discontinuation, as new ribosomes catch up with normal CYP3A4 turnover. Markowitz et al. 2003 demonstrated 1.66-fold increased oral midazolam clearance after 14-day SJW dosing, the canonical CYP3A4 probe response. Critically, Mai et al. 2003 showed that the hyperforin content of the preparation drives the magnitude — low-hyperforin extracts (e.g., Ze 117) produced minimal cyclosporine effect, conventional high-hyperforin extracts produced 45% AUC reduction. This matters because SJW preparations on the supplement market are unregulated and vary 5- to 10-fold in hyperforin content.

Evidence

Statins. Grapefruit raises CYP3A4-metabolised statins (simvastatin, lovastatin, atorvastatin) and does not raise CYP3A4-independent statins (pravastatin, rosuvastatin, fluvastatin). Lilja et al. 1998: 200 mL grapefruit juice three times daily for three days increased simvastatin AUC by ~16-fold (active acid metabolite +7-fold). Lilja et al. 1999: atorvastatin AUC +83% under similar conditions; pravastatin unchanged. The downstream signal is rhabdomyolysis: case reports cluster around simvastatin and lovastatin co-consumed with chronic grapefruit. SJW induces statin metabolism in the opposite direction — simvastatin AUC reduced ~62% by 14-day SJW co-administration Mannel 2004.

Immunosuppressants — the highest-stakes category. Cyclosporine and tacrolimus have narrow therapeutic windows (cyclosporine target 100–300 ng/mL trough), are heavily CYP3A4-dependent, and also rely on P-glycoprotein. Ruschitzka et al. 2000 reported two cases of acute heart transplant rejection after patients began self-medicating with St. John's wort; cyclosporine trough levels fell 50% within weeks. Hojo et al. 2011 and Mai et al. 2003 quantified the cyclosporine AUC drop and confirmed hyperforin-dose-dependence. Conversely, grapefruit can elevate cyclosporine/tacrolimus; though some early thinking explored grapefruit as an intentional cyclosporine-sparing strategy, the variable magnitude of effect made it clinically unworkable.

Calcium-channel blockers. Felodipine is the canonical case: Bailey et al. 1991 and many replications show 2–3× AUC increase from one glass of grapefruit juice, with measurable hypotension and reflex tachycardia. Lundahl et al. 1995 demonstrated the effect persists when grapefruit is consumed 24 h before felodipine — confirming the irreversible inhibition mechanism. Lundahl et al. 1997 showed the inhibition is maximal after the first glass — chronic exposure adds no additional inhibition, but a single intake is enough. Nifedipine, nisoldipine, and amlodipine are also CYP3A4-dependent though less affected than felodipine. SJW induction drops nifedipine plasma levels meaningfully though clinical consequences are less dramatic than for transplant drugs.

Hormonal contraceptives. Ethinylestradiol and progestins (norethindrone, levonorgestrel, desogestrel) are CYP3A4 substrates. Hall et al. 2003: 14-day SJW dosing reduced norethindrone AUC by 13% and increased ethinylestradiol clearance, with breakthrough bleeding in 7/12 women (vs 2/12 placebo) — a clinical signal of declining contraceptive coverage. Murphy et al. 2005 extended this: SJW increased the proportion of cycles with follicular growth (a marker of escaped ovulation suppression) and breakthrough bleeding. Schwarz et al. 2003 documented case reports of unintended pregnancy under SJW co-medication despite combined oral contraception. Grapefruit's contribution here is mostly theoretical/minor (mild AUC increase, no clinical signal of harm).

Psychotropics. Many CYP3A4-metabolised: midazolam, alprazolam, buspirone, quetiapine, aripiprazole, ziprasidone, trazodone, sertraline (partially). Grapefruit and SJW shift their plasma concentrations in opposite directions. Midazolam is the standard CYP3A4 probe in pharmacology: Greenblatt et al. 2003 documented near-doubling of oral midazolam AUC after grapefruit; Markowitz et al. 2003 documented 1.66-fold clearance increase after SJW. Serotonin syndrome is the special hazard when SJW (which has mild SSRI-like activity itself; see Linde et al. 2008) is combined with serotonergic antidepressants — a pharmacodynamic interaction added on top of the CYP3A4 induction.

Other categories worth naming. Piscitelli et al. 2000: SJW reduced indinavir (HIV protease inhibitor) AUC by 57% — the report that triggered FDA contraindication for SJW + antiretrovirals. Johne et al. 1999: SJW reduced digoxin AUC by 25% via P-gp induction (digoxin isn't a CYP3A4 substrate but the P-gp pathway runs in parallel). Schwarz et al. 2005: talinolol AUC reduced 31% by SJW, isolated P-gp effect.

Contraindications

The contraindication landscape is about the medication, not the food/herb — grapefruit and SJW are unsafe only in the context of a vulnerable co-medication. The major substrate classes the literature documents Bailey 2013 FDA 2021:

• Statins: simvastatin, lovastatin, atorvastatin (NOT pravastatin, rosuvastatin, fluvastatin)
• Calcium-channel blockers: felodipine, nifedipine, nisoldipine, amlodipine (less so)
• Immunosuppressants: cyclosporine, tacrolimus, sirolimus, everolimus
• Antiarrhythmics: amiodarone, dronedarone, quinidine
• Anticoagulants: apixaban, rivaroxaban (DOACs); warfarin (PD/PK signal)
• Antidepressants/anxiolytics: sertraline, citalopram, escitalopram, alprazolam, buspirone, quetiapine
• Hormonal contraceptives: ethinylestradiol-containing combined pills, progestin-only pills
• Antiretrovirals: indinavir, saquinavir, ritonavir, the integrase-inhibitor class
• Anti-cancer kinase inhibitors: imatinib, nilotinib, dasatinib, sunitinib, sorafenib (narrow therapeutic windows + CYP3A4 dependence)
• Erectile-dysfunction drugs: sildenafil, tadalafil (additive hypotension risk with grapefruit)
• Other narrow-window CYP3A4 substrates: colchicine (rhabdomyolysis), carbamazepine

Pregnancy and breastfeeding are independently relevant: SJW is contraindicated in pregnancy (potential abortifacient activity from hyperforin; teratogenicity data inadequate); grapefruit is not. The drug-interaction contraindications above apply across both sexes and all ages.

Protocol

The practical guidance synthesises across primary-care, pharmacy, and transplant-clinic consensus:

• Grapefruit timing. Because the inhibition is mechanism-based, spacing the food and the drug by hours does not work. Lundahl et al. 1995 demonstrated felodipine elevation persists 24 h after grapefruit; Greenblatt et al. 2003 documented 50% recovery in 24 h and full recovery only at ~72 h. The pharmacological default for affected drugs is total avoidance during therapy; the "three-day rule" (avoid for 72 h before any dose of the drug) is the practical maximum if the patient wants to eat grapefruit occasionally.
• Dose-response is shallow. Lundahl et al. 1997 showed maximal inhibition after the first 8-oz glass; bigger or more frequent intake adds little additional effect. There is no "safe small amount" — half a glass produces most of the inhibition.
• SJW washout. CYP3A4 induction decays over 1–2 weeks after discontinuation. For perioperative or peri-transplant management, the pharmacological default is to discontinue SJW ≥2 weeks before the dose-sensitive drug is started or adjusted.
• Substitute within drug class. Most affected drugs have a CYP3A4-independent alternative: pravastatin/rosuvastatin for simvastatin; loratadine for older sedating antihistamines; SNRI alternatives where psychiatric care permits. The choice belongs to the prescribing clinician.

Misconceptions

• "Only juice matters." Whole grapefruit, grapefruit segments, and grapefruit zest all carry furanocoumarins. The juice studies dominate the literature because juice standardises the dose, not because juice is uniquely active.
• "Only grapefruit." Seville (sour) orange, pomelo, and tangelo carry comparable furanocoumarin loads and produce similar inhibition. Sweet orange (Valencia, navel) does not — it lacks the relevant furanocoumarins. Lemons and limes are mild but not negligible (especially lime juice).
• "Take the drug at a different time of day." Doesn't work — see Protocol above.
• "Natural means safe." The most consequential herb-drug interaction in transplant medicine — acute organ rejection — was caused by an over-the-counter herbal supplement Ruschitzka et al. 2000. The "natural" framing is the failure mode (see below).
• "SJW dose matters more than brand." Hyperforin content varies 5- to 10-fold across products labelled with the same "300 mg standardised" extract Mai et al. 2003. The dose on the bottle does not reliably predict the induction magnitude.
• "All statins are equivalent for grapefruit." Pravastatin and rosuvastatin are CYP3A4-independent and unaffected by grapefruit.

Failure modes

• Self-medication concealment. Patients consistently fail to disclose herbal supplements to clinicians — the supplement is mentally categorised as "food" or "vitamins," not "medication." Henderson et al. 2002 documented this across spontaneous-reporting databases. The interaction surfaces as unexplained loss of efficacy: rising tacrolimus dose requirements, breakthrough seizures on carbamazepine, breakthrough bleeding on the pill.
• Pharmacist gap. Pharmacy interaction-checking software flags grapefruit on the patient information leaflet but rarely flags herbal products unless the patient enters them.
• "I only had a little." The dose-response is so shallow Lundahl et al. 1997 that intuition about "moderate consumption" is wrong. A single small intake matters; large intake does not matter more.
• Stopping SJW abruptly. Induction decays over 1–2 weeks; abrupt discontinuation in a steady-state tacrolimus patient causes rising tacrolimus levels as the induced enzyme returns to baseline. The opposite problem to the starting case but equally important.

Audience

By substrate-drug exposure:

• Solid-organ transplant recipients — highest stakes. Acute rejection on SJW is a documented fatal outcome Ruschitzka et al. 2000.
• Older adults on combination cardiovascular regimens — statin + calcium-channel blocker + DOAC together compound grapefruit risk; this population is disproportionately prescribed these classes.
• Women on hormonal contraception — SJW for premenstrual mood symptoms or depression is a common combination with contraceptive failure as a documented consequence Schwarz et al. 2003.
• People with HIV on antiretroviral therapy — the original index population for FDA SJW warnings Piscitelli et al. 2000.
• Patients on psychiatric medication — both for the CYP3A4 substrate interaction and for additive serotonergic effects from SJW.
• Cancer patients on oral kinase inhibitors — narrow therapeutic windows, CYP3A4-dependent, increasingly prevalent outpatient regimen.

Alternatives

Most affected drug classes contain at least one CYP3A4-independent member; substitution is the clinical default when grapefruit avoidance is impractical or SJW is desired:

• Statins → pravastatin, rosuvastatin, fluvastatin (CYP3A4-independent)
• Calcium-channel blockers → diltiazem, verapamil (less affected; though they have their own CYP3A4 inhibitory activity)
• Benzodiazepines → lorazepam, oxazepam, temazepam (glucuronidated, not CYP3A4)
• SSRIs → fluoxetine, paroxetine (different metabolism profile)
• Hormonal contraception under SJW → non-oral / non-systemic routes (copper IUD, depot progestin) are not rescued by SJW induction at the same magnitude

Substitution decisions are clinician-side, not patient-side; the article will frame the patient action as raise the question rather than change the prescription.

Credibility range

Optimist case (for the warning). The grapefruit and SJW interactions are some of the best-characterised drug-food/drug-herb interactions in clinical pharmacology. Mechanism is settled at the molecular level (furanocoumarins forming covalent adducts with CYP3A4; hyperforin activating PXR). Multiple human pharmacokinetic trials, with dose-response data and named-substrate clearance numbers, support the magnitude estimates. Case-series evidence covers the rare clinical consequences (rhabdomyolysis, transplant rejection, contraceptive failure). FDA labelling, transplant-society guidelines, and pharmacology textbooks converge. The intervention — avoidance, or substitution within drug class — is cheap, reversible, and low-effort.

Skeptic case. Most healthy adults take none of the affected drugs and the warning is irrelevant to them. For patients on affected drugs, the clinically reported events (rhabdomyolysis, rejection, breakthrough bleeding) are case reports, not rate estimates — population-level harm magnitude is not well quantified. The grapefruit-statin signal is dose-dependent and may be modest for low-dose simvastatin in occasional grapefruit consumers. The "three-day rule" is a conservative pharmacology heuristic, not a tested clinical recommendation. SJW preparations vary 5- to 10-fold in hyperforin content, so blanket "SJW is dangerous" warnings overgeneralise; low-hyperforin preparations cause minimal interaction Mai et al. 2003. Some interaction warnings on patient leaflets are theoretical rather than clinically documented.

Author's call. The mechanism, magnitude, and direction of these interactions are not in dispute — this is settled CYP3A4 pharmacology with consistent replication. The honest framing for a reference catalogue is: high confidence in the existence and direction of the interactions; less confidence in the rate of catastrophic outcomes at the population level; very high confidence that affected-population harms (rejection, rhabdomyolysis, pregnancy) are real and avoidable. The right reader action is awareness and disclosure to a prescribing clinician — not panic about a single grapefruit slice. Evidence score 5 (settled mechanism, multiple RCTs); controversy score 1 (minor ongoing debate about hyperforin-content cutoffs and clinical magnitude in low-risk substrate use).

Stakeholder + incentive map

• Pharmacovigilance / regulatory: FDA, EMA, MHRA. Strong push to label CYP3A4 substrates with grapefruit and SJW warnings; the labelling is uniform and conservative.
• Transplant clinicians: The community most attuned to the SJW interaction (and to grapefruit avoidance for tacrolimus/cyclosporine). Universally counsel avoidance of both.
• Herbal supplement industry: Commercial incentive to downplay interaction risk. SJW is sold as a "natural antidepressant" with disclaimers that vary by jurisdiction; in Germany SJW is a prescription pharmaceutical, in the US it is OTC. The Cochrane SJW review Linde et al. 2008 finds the herb has modest antidepressant efficacy comparable to SSRIs in mild-moderate depression — this is real, and not in conflict with the interaction warning.
• Grapefruit growers / juice industry: Smaller commercial pushback; industry-funded work has explored low-furanocoumarin grapefruit cultivars and processing methods.
• Pharmacology academic community: The Bailey group at Western Ontario is the dominant source for grapefruit-interaction research; SJW research is more distributed.
• Pharmacists at point-of-dispensing: The recurring backstop. Grapefruit warnings reliably surface on the patient leaflet; SJW warnings depend on the patient mentioning the supplement.

Population variability

• Baseline CYP3A4 expression varies ~10-fold between individuals. Patients with lower baseline intestinal CYP3A4 (genetic, age-related, or already pharmacologically inhibited) experience proportionally larger relative inhibition from grapefruit. Pharmacogenomic CYP3A4/5 polymorphisms add another layer of variability.
• Older adults have higher exposure on average because they take more CYP3A4 substrates (statins, calcium-channel blockers, DOACs) and have lower baseline CYP3A4 reserve.
• Women on combined hormonal contraception have CYP3A4 baseline depressed by ethinylestradiol, raising relative sensitivity to grapefruit interactions with co-medications; the same baseline depression makes ethinylestradiol itself more sensitive to SJW induction.
• Pediatric population — interactions extrapolated from adult data; less direct evidence. Less commonly an issue because affected drug classes are less prescribed.
• Severe hepatic impairment alters CYP3A4 function unpredictably; effect of grapefruit/SJW is harder to predict.

Knowledge gaps

The most important uncertainty is population-level event rates: how often does grapefruit consumption in real-world statin users cause clinically meaningful rhabdomyolysis? How often does SJW co-medication cause contraceptive failure in real-world OCP users? Pharmacokinetic data are abundant; outcome-incidence data are sparse and come mostly from case reports and pharmacovigilance signals. A second gap is whether low-hyperforin SJW preparations (such as Ze 117) are safe enough to be the default — preliminary evidence says yes but the supplement market does not distinguish them. A third is whether furanocoumarin-free grapefruit cultivars can be commercialised at scale; the Paine et al. 2006 paper proved the concept but commercial uptake is minimal. Finally, the interaction profile of newer kinase inhibitors and oral oncology drugs is being mapped in parallel with their approval — a moving target rather than a settled domain.

Scope. The brief named the substance (CYP3A4 modulation by grapefruit and St John's wort) and five drug classes (statins, immunosuppressants, calcium-channel blockers, hormonal contraceptives, psychotropics). The article covers all five end-to-end in the evidence section. Additional affected classes (DOACs, warfarin, antiretrovirals, ED drugs, amiodarone) get a brief signpost rather than full treatment — they would dilute the five-class story and most are clinician-managed rather than patient-action-relevant.

One entry, two substances. Grapefruit and St John's wort have nothing in common chemically or culturally; pairing them in one entry is justified only by the shared mechanism (CYP3A4 modulation). The risk of doing this is that the entry feels like two articles awkwardly fused. Mitigation: the mechanism section frames them as mirror images from the start, and every drug-class subsection in §evidence handles both directions. If this entry ever needs to split, the natural cut is by substance (grapefruit-CYP3A4-inhibition, st-johns-wort-CYP3A4-induction), not by drug class.

Rating difficulty: short-term health. Scored 3 on health_short_term, which is high for an entry whose action is "avoid two specific dietary items." The score reflects the substance's protective magnitude in the at-risk population (someone on simvastatin avoiding rhabdomyolysis is a meaningful near-term health win), not the broad reader for whom the entry is informational only. Alternative was scoring 1 or 2 to reflect the small affected fraction; rejected because the rating framework says score the substance, not the article's coverage. Same logic for longevity at 2.

Zero scores on energy / focus / sleep / mood / beauty. No direct effect on any of these from either substance considered as a CYP3A4 modulator. St John's wort has its own mood effect when used as an antidepressant, but that's a different entry — including the mood effect here would conflate the interaction story with the treatment story. Flagged in editor notes as a future-link.

Category. Placed under medical rather than food or supplements. The entry's centre of gravity is pharmacology — a drug interaction warning — not the nutritional profile of grapefruit or the herbal-supplement landscape. A reader looking for grapefruit-as-food or St John's wort-as-antidepressant should find separate future entries.

Action verb. Chose avoid over know. know would have been defensible (the entry is informational for the majority of readers) but avoid matches the substantive action the affected subgroup needs to take, and the article's protocol section makes that the centre of the recommendation.

Future-link candidates.

• st-johns-wort-for-depression — the same herb, evaluated as a mild-to-moderate depression treatment. The Cochrane evidence (Linde et al. 2008) is genuinely positive; the topic deserves its own entry rather than a footnote here.
• polypharmacy-and-supplement-disclosure — the generalised principle: every herbal supplement, every dietary change deserves a check against the prescription list. Grapefruit and St John's wort are the most-cited examples, not the only ones.
• simvastatin-vs-rosuvastatin or a broader statin-choice entry — the CYP3A4-substrate distinction is one of several axes on which statins differ.
• transplant-medication-basics — narrow audience but very high stakes; this entry signposts the issue but doesn't carry full transplant-pharmacy detail.

Hard decision: the rhabdomyolysis numbers. The 16-fold simvastatin AUC figure from Lilja et al. 1998 is the most-cited number in the grapefruit literature but came from an extreme dosing protocol (200 mL double-strength juice three times daily for three days) that doesn't match real-world consumption. Reported the number but contextualised it as the worst-case ceiling rather than the typical exposure. The intermediate exposures (one glass at breakfast on a stable statin dose) have less rigorous data and the article doesn't pretend otherwise.

Voice tension. The substance is technical (cytochrome P450 enzymology) and the reader is mostly a layperson trying to figure out if their breakfast is a problem. Tried to keep the mechanism section in plain-language analogies (enzyme "chewing up" the drug) and push the trial-level detail into science callouts. The friend-test bar is harder here than on most entries because the topic naturally pulls toward textbook voice.